Content uploaded by Jeroen Louwe Kooijmans
Author content
All content in this area was uploaded by Jeroen Louwe Kooijmans on Oct 12, 2018
Content may be subject to copyright.
Setting Sail in Scandinavia
An analysis of the evidence and arguments
By
Jeroen Louwe Kooijmans
Cover illustration: Photo by author
MASTER THESIS
Submitted in partial fulfilment of the requirements
for the degree of Master of Arts
in Maritime Archaeology
Jeroen Louwe Kooijmans
Department of History
University of Southern Denmark
Esbjerg, Denmark
17 September 2018
Supervisor:
Thijs J. Maarleveld
Absence of evidence is
not evidence of absence
- Unknown
i
Acknowledgements
First of all I would like to thank my parents who made it possible for me to fully focus on my
studies throughout the years. Furthermore, I would like to thank my father for also correcting
my English. My supervisor Thijs Maarleveld has given useful advice throughout the writing
and thinking process and always made time for a conversation. I would sincerely like to thank
Thijs for this and the many interesting lectures he has given throughout my master. Finally, I
would like to thank Heidi E. Vink for her support and for her critical comments when needed
and when not. I am glad that our lives have been brought together by our common interest in
maritime archaeology and ships.
ii
Abstract
This thesis concerns the discussion regarding the adoption of the sail in Scandinavia. It is the
aim of this thesis to assess the prevailing theory that the sail was a late adoption and to
provide an overview for others. Firstly the problems regarding the inductive method within
archaeology are discussed. Secondly the evidence is presented that is used within the
discourse. Aspects that may affect their validity as premises are presented. Thirdly the
arguments are presented, discussed and their strength assessed. The assessment of the
arguments is based on the validity of the premise and the relevance of the premise for the
theory. Since archaeology is mainly a discipline based on interpretation this assessment is
unavoidably subjective. A clear methodology is therefore presented for the assessment, for
which clearly defined qualitative scales are used. When the author´s own subjective
judgement is used then it will be made apparent. It is thus hoped that this thesis will provide
an overview of the evidence and make it possible for others to assess the strength of the
prevailing theory without having to rely on assumptions made by others.
It is concluded that there has been a lack of discussion regarding the representativeness of the
evidence, that the constructional prerequisites for a vessel to carry a sail which have excluded
an earlier use are not valid and that there are no reasons to exclude the possibility that the sail
was used at an earlier date than currently thought.
iii
Danish abstract
Afhandlingen angår diskussionen omkring anvendelse af sejlet i Skandinavien. Formålet med
afhandlingen er at vurdere den nuværende teori at seilet kom sent i brug og at give et overblik
for andre. For det første diskuteres problemerne med den induktive metoden inden for
arkæologi. For det andet fremlægges beviser der anvendes inden for diskursen. Aspekter som
kan påvirke deres gyldighed som præmiser præsenteres. For det tredje er argumenterne
præsenteret, diskuteret og deres styrke vurderet. Vurderingen af argumenterne er baseret på
gyldigheden av præmissen og relevansen af den for teorien. Da arkæologi hovedsageligt er et
felt baseret på fortolkning, er denne vurdering uundgåelig subjektiv. Der gives derfor en klar
metode til vurderingen, for hvilke der anvendes klart definerede kvalitative skalaer. Når
forfatterens egen subjektive dømmekraft bruges, vil dette blive gjort klart. Det håber således,
at denne afhandling vil give et overblik over beviserne og gøre det muligt for andre at vurdere
styrken af den nuværende teori uden at skulle stole på antagelser fra andre.
Det konkluderes, at der har været manglende diskussion vedrørende bevisets
repræsentativitet, at de konstruktionsmæssige forudsætninger for at sejlet et fartøj der har blitt
brugt for at udelukket en tidligere anvendelse av sejlet ikke er gyldige, og at der ikke er nogen
grund til at udelukke muligheden for at sejl blev brugt tidligere end det er for tiden troet.
Table of Contents
Acknowledgements ..................................................................................................................... i
Abstract ...................................................................................................................................... ii
Danish abstract ......................................................................................................................... iii
List of Figures .......................................................................................................................... iv
Chapter 1. Introduction ............................................................................................................. 1
1.1 The sail in Scandinavia ............................................................................................................... 1
1.2 Incentive ....................................................................................................................................... 1
1.3 The aims ....................................................................................................................................... 1
1.4 Methodology ................................................................................................................................ 2
1.5 Terminology ................................................................................................................................. 2
1.6 Layout .......................................................................................................................................... 2
Chapter 2. Theory ...................................................................................................................... 3
2.1 On the problems of induction within archaeology ................................................................... 3
2.2 Argumentation analysis .............................................................................................................. 4
Chapter 3. Methodology ............................................................................................................ 5
3.1 Structure of the arguments ........................................................................................................ 5
3.2 Assessment of an argument ........................................................................................................ 6
3.3 Assessment of the probability and relevance ............................................................................ 7
3.3.1 Vessel remains ...................................................................................................................................... 7
3.3.2 Iconography .......................................................................................................................................... 7
3.3.3 Textual evidence ................................................................................................................................... 7
Chapter 4. The evidence ............................................................................................................ 9
4.1 Vessel remains ............................................................................................................................. 9
4.1.1 Hjortspring ........................................................................................................................................... 9
4.1.2 The Nydam vessels ............................................................................................................................. 14
4.1.3 The Kvalsund vessels ......................................................................................................................... 22
4.2 Iconography ............................................................................................................................... 26
4.2.1 The Gotlandic picture stones .............................................................................................................. 26
4.2.2 The Häggeby stone ............................................................................................................................. 32
4.2.3 The Karlby pebble .............................................................................................................................. 33
4.2.4 The Weser Bone ................................................................................................................................. 35
4.2.5 The Hvirring razor .............................................................................................................................. 35
4.2.6 The vendsyssel vessel ......................................................................................................................... 36
4.2.7 Rock art vessels with masts and/or sails ............................................................................................. 36
4.3 Textual evidence ........................................................................................................................ 40
4.3.1 Tacitus´ Germania .............................................................................................................................. 40
4.3.2 Gildas´ De Excidio Britanniae ............................................................................................................ 41
4.3.3 Sidonius Appolinaris´ Letters ............................................................................................................. 42
4.3.5 Procopius´ History of the Wars .......................................................................................................... 44
Chapter 5. The Arguments ...................................................................................................... 45
5.1 The vessel remains .................................................................................................................... 45
5.1.1 The Arguments ................................................................................................................................... 45
5.1.2 What is looked for? ............................................................................................................................. 45
5.1.5 The evolutionistic view....................................................................................................................... 49
5.1.7 Assessment of the arguments .............................................................................................................. 53
5.2 Iconography ............................................................................................................................... 55
5.2.1 Bronze/ early Iron Age iconography .................................................................................................. 55
5.2.2 The Karlby pebble .............................................................................................................................. 56
5.2.3 The Weser Bone ................................................................................................................................. 56
5.2.4 The Gotlandic Picture Stones ............................................................................................................. 57
5.3 Historical evidence .................................................................................................................... 59
5.3.1 Tacitus ................................................................................................................................................ 59
5.3.2 Gildas .................................................................................................................................................. 60
5.3.3 Sidonius .............................................................................................................................................. 60
5.3.4 Procopius ............................................................................................................................................ 62
5.4 Assessment of the prevailing theory ........................................................................................ 63
Chapter 6. Discussion .............................................................................................................. 64
Chapter 7. Conclusion ............................................................................................................. 67
Glossary ................................................................................................................................... 70
Bibliography ............................................................................................................................ 72
iv
List of Figures
Figure 1. Table for determining the strength of an argument. After Reitan, 2009, p. 133......... 6
Figure 2. Hjortspring during excavation. After Rieck & Crumlin-Pedersen, 1988, p. 58. ........ 9
Figure 3. Cross-section of Hjortspring after Crumlin-Pedersen, 2003, p. 30. ......................... 10
Figure 4. Reconstruction and lines plan of Hjortspring by Johannessen. After Crumlin-
Pedersen, 2003, p. 36. ............................................................................................. 11
Figure 5. The trussing rope fastened to the cleats at the stems. After Hocker, 2003, p. 85. .... 12
Figure 6. The preserved parts of Nydam B. After Åkerlund, 1963, p. 51 ................................ 14
Figure 7. Drawing made by Engelhardt of the keel of Nydam C. After (Engelhardt, 1865, p.
11). ........................................................................................................................... 15
Figure 8. The planks recorded from the Nydam A boat. After Engelhardt 1865, figure IV. ... 15
Figure 9. Drawing of Nydam B stem scarf. After Shetelig & Johannessen, 1930, p. 12,
redrawn by author. ................................................................................................... 16
Figure 10. The different reconstructions made of the cross-section amidships. After Åkerlund,
1963, p. 36. .............................................................................................................. 17
Figure 11. The longitudinal construction as imagined by Åkerlund. After Åkerlund, 1963,
figure IV. ................................................................................................................. 18
Figure 12. Keel section of the Nydam pine boat. After Engelhardt 1865, p. 11. ..................... 18
Figure 13. How Åkerlund imagined the ends were constructed. After Åkerlund, 1963, p. 99.
................................................................................................................................. 19
Figure 14. Reconstruction drawing of Kvalsund II by Johannessen. After Shetelig &
Johannessen, 1929. .................................................................................................. 22
Figure 15. Showing the typology in shape and their respective dating by Lindqvist. After
Nylén & Lamm, 2003, p. 172. ................................................................................. 26
Figure 16. A type A stone with a representation of a ship. The Bro I picture stone. After
Lindqvist, 1941. ....................................................................................................... 26
Figure 17. Type B picture stones. From left to right: Rikvide, Fole K & Endre Skog. After
Lindqvist, 1941. ....................................................................................................... 27
Figure 18. The Lärbo St. Hammars I picture stone to the left and the Hejnum Riddare stone to
the right. After Lindqvist, 1941. .............................................................................. 27
Figure 19. The overlap of Varenius and Lindqvist´s divisions. After Varenius, 1992, p. 56. . 29
Figure 20. The divisions made by Varenius and Lindqvist with their respective date. Made by
author. ...................................................................................................................... 29
Figure 21. The bronze brooches referred to by Lindqvist. After Nerman, n.d., fig. 81 & 87. . 29
Figure 22. Gotlandic picture stones in Salin style II and with similar band decoration. From
top left to bottom right: Sanda Sandegårda I, Fole K, Endre Skog, Stenkyrka IX,
Lärbo Norder-Ire I and Hellvi St. Ire III. After Lindqvist, 1941. ............................ 30
List of Figures v
Figure 23. The bronze fibulae referred to by Lindqvist. After Nerman, n.d., fig. 23, 24 & 27.
................................................................................................................................. 30
Figure 24. The Häggeby stone. After Montelius, 1906, p. 197. ............................................... 32
Figure 25. Photo by Djurslands Museum. After Rieck & Crumlin-Pedersen, 1988, p. 129. ... 33
Figure 26. Drawning by Werner Karrasch. After Rieck & Crumlin-Pedersen, 1988, p. 133. . 33
Figure 27. Drawing by Hans Drake. After Westerdahl, 2015, p. 43. ....................................... 33
Figure 28. Examples of Dorestad coins. After Lebecq, 1990, p. 88. ....................................... 34
Figure 29. A Hebey coin with on one side a ship and on the other a horned animal. After
Arbman, 1940, Taf. 142, 9. ..................................................................................... 34
Figure 30. The Weser ship carving. After Göttlicher, 1989. .................................................... 35
Figure 31. The Hvirring razor. As published by Müller (left) and Worsaae (right). After
Müller, 1897, p. 444 and Worsaae, 1859, p. 36. ..................................................... 35
Figure 32. The carving of the vendsyssel vessel. After Bengtsson, 2017, p. 53. ..................... 36
Figure 33. The carving from Hafslund III. After Marstrander, 1963, Pl. 59............................ 37
Figure 34. Askum-44. To the left a photo and to the right the interpretation made. After
Humbla & von Post, 1937, p. 71. ............................................................................ 38
Figure 35. Drawing of Askum-44 as interpreted by Boel Bengtsson. After Bengtsson, 2017, p.
67. ............................................................................................................................ 38
Figure 36. An overview of several depictions with a possible mast and/or sail. ..................... 39
1
Chapter 1. Introduction
1.1 The sail in Scandinavia
This thesis concerns the question when the sail was first adopted in Scandinavia. The earliest
widely accepted archaeological evidence are depictions of sailing vessels on the Gotlandic
pictures stones from the 6th and 7th centuries AD. The earliest find of a sailing vessel is
currently the Oseberg ship from the start of the 9th century. This is much later than the first
evidence of sailing vessels in the more southern parts of north-western Europe. Here the sail
is believed to have been used from at least the first century BC (McGrail, 2001, p. 211). The
estimates for the first use of the sail in Scandinavia have varied from the 6th to the 8th century
AD. This variation is mainly due to the dates given to the picture stones. It is also thought that
the change from paddling to rowing to eventually sailing vessels is also visible in the
development of the constructional features of the pre-historic vessels. Thus Hjortspring from
the 4th century BC is a slender vessel for calmer waters which was propelled by paddles.
Nydam B from the 4th century AD is propelled by oars and is built sturdier for rougher waters
but still has a keel plank and is therefore not suited for sailing. From the end of the 7th century
there is the Kvalsund vessel which has a ´proper´ keel and a hull appropriate for sailing, but
there were no signs of a mast-step, mast or rigging. The Oseberg ship is the first sailing vessel
from Scandinavia although its keelson and mast partner are too small for its size showing that
the sailing vessel was not yet fully developed. Hundreds of years of experimentation finally
resulted in an advanced sailing ship, the Gokstad ship from the end of the 9th century
1.2 Incentive
The development towards sailing in Scandinavia sketched above is the theory that currently
prevails, here on out referred to as ´the prevailing theory´. During my bachelor it was pointed
out to me that the sail was curiously late adopted in Scandinavia. The main question that
remained to be answered was; Why? This question remained with me and was the incentive
for this thesis. During the reading of the literature in preparation for this thesis many issues
became apparent, resulting in the current question; Was it? To me it seems necessary to
provide a thorough analysis since many of the problems connected to inductive reasoning and
the archaeological record, that normally should be taken into account in every archaeological
discussion, have hardly been discussed in the discourse regarding the adoption of the sail in
Scandinavia. Many of the arguments are based on underlying assumptions which have been
too readily accepted.
1.3 The aims
It is not the aim of this thesis to give an answer to the question when the sail was adopted, but
rather to assess the current basis for the prevailing theory that it was a late one.
Currently there is no literature that provides a complete overview of the evidence and the
arguments that are of relevance for this theory. When the adoption of the sail is discussed
only a selection of the evidence is given, resulting in a one-sided view. An additional aim is
therefore to provide an overview of the current evidence and arguments to enable others to
make up their own opinion regarding the use of the sail in the North, rather than having to
Chapter 1. Introduction 2
accept someone else’s opinion on the basis of only a partial and possibly biased overview of
the evidence.
Another ´issue´ with the relevant literature is that most of it is written in a Scandinavian
language. In particular the literature that makes it possible to verify the conclusions by other
authors, e.g. excavation reports. Those who cannot read such texts are dependent on the
general overviews written in English. These overviews are often only a summary of the
conclusions drawn in the Scandinavian discourse. That is why I have chosen for a more in-
depth description of the evidence, especially of those aspects where there is room for doubt
and which may point to different interpretations.
1.4 Methodology
In order to assess the probability of the prevailing theory and to provide an overview, the
evidence that has been presented in the discourse will be presented here. Information that is of
relevance for the assessment of the theory will be included. In order to assess the theory the
arguments that until now have been formulated that are directly based on the evidence will be
presented, their structures made visible and their strength assessed. By assessing each
argument separately, the fallacies and underlying assumptions will become apparent. It is
chosen to restrict the assessment to the arguments that are based directly on the evidence.
Arguments which are based on the probability of other theories is too wide of a subject for
one thesis.
1.5 Terminology
This thesis regards the adoption of the sail in ´Scandinavia´. This term will also be used
throughout this thesis since it has been the preferred term within the discourse. Using such a
modern construct is however not without its issues. The idea of Scandinavia as being a
separate region in pre-history, not only geographically but also culturally, is often too readily
accepted as will also become apparent in this thesis.
The term ´sail´ within this thesis is defined as any intentional method of using a piece of
material to catch the wind and thereby propelling a vessel. When the verb ´sailing´ is used, the
propulsion of a vessel through the use of a sail is meant.
Definitions regarding the nautical terminology can be found in the glossary.
1.6 Layout
In chapter two the problems regarding method of induction are discussed and the theory of the
argumentation analysis presented. In chapter three the methodology of the argumentation
analysis is presented. Here a further explanation of how the argument analysis and assessment
will be conducted is given. In chapter four the evidence is presented with a focus on
information that is of relevance for the assessment of the arguments. The main arguments are
presented in chapter five. The probability of the premises and their relevance for the theory of
a late adoption will be assessed. Chapter six is a discussion of the main problems with the
current discourse. Chapter seven is the final conclusion where the main uncertainties of the
current evidence are pointed out.
3
Chapter 2. Theory
2.1 On the problems of induction within archaeology
In order to deliver a proper analysis of the argumentation, the problems regarding the method
of induction should be taken into account. The primary method used in archaeology is that of
induction, i.e. inferring theories from the archaeological material. The major concern with the
inductive method is that a theory inferred from the archaeological material will always remain
no more than a probability. It can always turn out to be false (Popper, 1968, p. 27). The crux
of archaeology is the interpretation of the material. It is the theory that is most likely to be
correct that will be accepted until a theory is put forward which is seen as more likely.
Interpreting or judging which theory is most likely is highly subjective. This inbuilt
judgement of what is deemed correct or true is dependent on the researcher´s background.
Where he/she grew up, what the point of view of their teachers and professors was, which
literature one has read, etc. Although one is taught to be critical and self-conscious of one’s
bias, it is impossible to completely avoid it.
Although this being the case, the possibilities for the interpretation of the archaeological
material is not unlimited. The material does provide us with boundaries. This is primarily the
case for the lower levels along the ladder of inference, that is, technology and subsistence
strategies (for the ladder of inference see Hawkes, 1954). Higher up the ladder it becomes
more prone to discussion and doubt. Conveniently, the usage of the sail belongs to the lowest
level, that of technology. Direct evidence for this in the form of for example a mast provides
us with a terminus ante quem to our interpretations. What is problematic is that the
archaeological material related to the usages of the sail does not and cannot provide us with a
terminus post quem. An earlier date will always remain open for discussion. It is for example
possible to falsify the prevailing theory when finding a mast from an earlier period, but the
absence of a mast in the archaeological record is not evidence for the absence of the sail.
Absence of evidence is not evidence of absence.
Popper´s (1968) method of falsification is partly applicable to archaeology. A hypothesis
holds until evidence to the contrary is provided. If we take as an anti-thesis; the usage of the
sail occurred earlier than the 8th century, it seems possible to apply the falsification method of
Popper. Criteria to identify a sailing vessel can be listed, if these are present in a vessel this
will work as a falsification of the prevailing theory. However, such criteria can be
problematic. What becomes apparent in the literature is that many presuppose that the absence
of a protruding keel is indicative of the absence of a sail. One can simply not make such an
assumption. The question to be answered prior to such a conclusion is; which criteria did the
boatbuilders and the intended user follow and choose for a sailing vessel? This question is
difficult to answer since it will lead to a problem of identification. In order to recognize a
sailing vessel in the archaeological record one needs to know the criteria, but in order to know
what these criteria were, one first needs to find a vessel which was used for sailing. This is
also applicable to the identification of a sail in iconography. In order to identify a figure of a
vessel as having a sail one needs to have a prior idea what one is looking for. One has to be
careful that this is not done with a preconception what a sail should look like (if not rested on
evidence) since this may result in a misinterpretation of the iconography. Clearly stating what
criteria or ´signs´ one is looking for is of high importance so that their validity may be
discussed. What is it one looks for to identify if a vessel was used for sailing? How does on
recognise a representation of a sail?
Chapter 2. Theory 4
The intended function of a ship is a determinant for the design criteria. Different functions
result in different shapes and features. A possible reason for using a flat keel instead of a
protruding keel is ease of beaching. The absence of a protruding keel will however result in
more leeway. That is the difficulty with designing a vessel, choosing which functions are
deemed more important than others. Compromises have to be made when designing a vessel,
one feature will always affect another. It is therefore always important to establish the
intended function of a ship when discussing its capabilities. Were the boatbuilders unaware of
a certain advantageous feature or was it a choice well-thought through?
The main problem with inductive reasoning is how representative the evidence is. In other
words, do the generalising theories have any validity? Especially within archaeology this is
one of the major concerns which needs to be taken into account. The remains that have been
left to us by the past are only a piece of it, of which even less is eventually recorded and
published (Maarleveld, 2010). The material that has been intentionally left in the ground, such
as grave goods, has also been selected prior to deposition, likely giving us a distorted view of
the actual situation. It is therefore always important to take into consideration the context of
the find, since these tend to indicate the type of vessel (Crumlin-Pedersen, 1990, p. 98). So, if
a vessel is found which does not have any indication of having had a sail, can it be stated that
there were no vessels with sails? Even if a hypothesis holds ground it is not redundant to
discuss other possibilities and point out the above concerns, especially with such a limited
source available.
What is especially relevant for the discussion of the sail is the question for which region the
material is representative. Although the Oseberg ship is the first conclusive evidence of the
use of sail in Scandinavia, can it be stated that everyone in ´Scandinavia´ was using the sail?
This argument can also be used in the other direction. If we accept that no sail was being used
based on the Nydam vessels (and thereby accepting that these ships are representative for all
of the vessels) does this mean that the sail wasn´t used in other regions than those from which
these vessels originated? Regional diversity has to be taken into account when discussing the
usage of the sail in ´Scandinavia´.
2.2 Argumentation analysis
What is problematic when assessing the arguments is that these are based on interpretations.
They cannot simply be determined to be false or true. It is therefore not possible to judge the
validity based on logical thought. It will always remain a subjective opinion. In this thesis it is
specified when the author´s judgement of the probability comes into play in order to ensure
that the reader can decide whether to agree or not. In this thesis the arguments will be
analysed and presented, cases of faulty logic will be criticised and other possibilities of
interpretations will be pointed out. Even though there is no direct evidence against a
hypothesis this does not mean that possible inadequacies should not be pointed out. It is
important to point out other possibilities and not to just focus on one. The choice for the most
probable theory remains a subjective opinion. For this thesis a demarcation for the
argumentations has to be set. The focus is therefore solely on arguments which are inferred
from the evidence discussed in chapter four. Several arguments in the discourse are based on
other theories, for example the size and composition of the migration to Britain or the
connectivity between Europe and Scandinavia (Haywood, 1991, 2006; Jones, 1996). A
thorough analysis of such arguments are too large of a subject and because these are not
directly based on the evidence found in Scandinavia they are often based on assumptions.
5
Chapter 3. Methodology
3.1 Structure of the arguments
An argument consists of several statements. These statements are commonly divided into
premises and conclusions. A premise is a statement for which another statement will argue it
to be a reason to accept a conclusion (Audi, 1999, p. 43). When analysing an argument it is
the validity of the premises and the relevance of these to the conclusion that are of
importance. It is therefore the aim to identify the structure of an argument, so that the
relevance can be judged. Within archaeology, due to it mainly being interpretation, it will
rarely be the case to find arguments which are logically valid (that is, if the premises are true
then the conclusion has to be true as well).
Since the arguments in archaeology are rarely logically valid they are susceptible to criticism.
In archaeology the premises on which theories are based are interpretations of the
archaeological material. The premises therefore have to be discussed thoroughly. An example
is judging whether a vessel was a sailing vessel or not. As mentioned, this can be quite
problematic. The relevance of the premise for the conclusion is where the concerns regarding
the representativeness, as discussed above, come into the picture. The validity of the
relevance of a premise depends mainly on the accepted theories in the scientific field.
The conclusions, the premises and the statements to support the relevance of these for the
conclusions, will be identified. Arguments will be noted using the following schematic:
Thesis
Premise
Relevance
For overviews the following schematic will be used:
T
T
Pn
P1
C1
Cn
PnP1
CnP1
T: Thesis
T: anti-Thesis
P: Premise
C: Counter-Premise
The thesis (T) is the prevailing theory that the usage of the sail was a late phenomenon in
Scandinavia whilst the anti-thesis (T) is the opposite of it i.e. that it was not a late
phenomenon. The thesis is supported by premises (P). The validity of the premise can be
further argued for or against (PnP1 and CnP1 respectively). Direct arguments for an earlier date
support the anti-thesis and are thereby counter premises (C) for thesis. The arguments for the
relevance of a premise will not be presented in this schematic but will be discussed and
evaluated in-text.
Chapter 3. Methodology 6
3.2 Assessment of an argument
The relevance of the premise for the conclusion is often based on a generalisation and can
always turn out to be false. An inductive argument can therefore never be true as with a
logical valid argument. A strong inductive argument is; if it is accepted that the premises are
true than it is probable that the conclusion is correct.
It is difficult to determine the strength of an argument since this cannot be done in a precise
quantitative manner. Due to the arguments being mainly based on interpretations it is
impossible to evaluate the argument without this evaluation being subjective. However, a
transparent and systematic methodology will be used in order to enable the reader to decide
whether to agree or not. The degree of probability will be described using qualitative scales
(After Reitan, 2009, pp. 130–133). For evaluating the probability of a premise the following
scale will be used:
- High: Determined to be true/ an established fact.
- Intermediate: reasonably probable/ sure.
- Weak: a minimum of preponderance.
The degree of relevance will be determined by the following scale:
- Strong: Given that the premise is true, then it is probable that the conclusion is correct.
- Intermediate: Given that the premise is true, then there is reasonable reason to believe
that the conclusion is correct.
- Weak: Given that the premise is true, then there is reason to believe that the
conclusion is correct, without giving it much persuasion.
The total strength of an argument can then be determined using the following chart:
Relevance
Strong
Intermediate
Weak
High
High
good
Weak
Probability
Intermediate
good
good
Weak
Weak
Weak
Weak
Weak
Figure 1. Table for determining the strength of an argument. After Reitan, 2009, p. 133.
This functions only as an indication since it is a subjective grading of the arguments. It is up
to the reader to decide whether to agree or not.
Chapter 3. Methodology 7
3.3 Assessment of the probability and relevance
Chapter four provides an overview of the evidence (i.e. premises) that is being used in the
discussion. For assessing the probability and relevance of a premise there are several aspects
that need to be focused upon. What these aspects are depends on the type of evidence. A
division has been made in; boat remains, iconography and textual evidence. Each of these
have their own interpretative issues. Certain information specific for each type of evidence is
therefore crucial for the assessment, especially regarding the uncertainties. The consequences
these issues have for the arguments will be discussed in chapter five.
3.3.1 Vessel remains
The probability of a premise based on vessel remains is mainly dependent on the probability
of their reconstruction (this includes their propulsion). In order to assess a reconstruction it is
important to discuss the excavation, the preservation and other proposed reconstruction(s).
Within the discourse there has been a focus on a few constructional criteria a vessel requires
to have in order for it to be sailed. These constructional features will be discussed in more
detail. For the assessment of the relevance of the vessel remains for the prevailing theory the
representativeness has to be considered. By doing so the validity of any generalising
statement may be assessed. It is therefore needed to discuss the find context, its date and
provenance. The function(s) of a vessel may be deduced from the find context whilst the date
and provenance or clearly of importance for the question surrounding representativeness.
3.3.2 Iconography
The strength of an argument based on iconography is often difficult to assess because it relies
on two layers of interpretation. The first level being the interpretation of which lines are
supposed to indicate a figure and the second one being the interpretation of the figure itself. It
has been attempted to make the former objective through different methods of documentation.
However the judgement which lines were supposed to be seen and the exact borders of these
lines remains somewhat subjective. The latter remains the most subjective; What is depicted?
This will likely remain an area of discussion and it is often not possible to determine with
certainty what is depicted. These difficulties are especially apparent in the interpretation of
rock art. Understanding what is depicted also encapsulates a further understanding beyond the
obvious. It is for example quite clear that the Gotlandic picture stones depict vessels, but what
further message are they supposed to convey? This needs to be considered in order to assess
whether any generalising conclusions drawn on the basis of the iconography have any
validity. Thus apart from the date and provenance, what the depiction is meant to display
should also be discussed.
3.3.3 Textual evidence
It is important to first establish what was actually written. For this I, and many others, are
dependent on the translation made by others. There are often several translations of the same
text, each giving a somewhat different nuance to the relevant sentences. Several times I have
therefore included different translations to point out the uncertainty.
It is often a single sentence that is taken as evidence. In order to avoid an incorrect
understanding the sentence needs to be seen in the context of the paragraph and the paragraph
Chapter 3. Methodology 8
in the context of the entire text. Then it may be possible to fully understand the message that
the author wanted to convey.
The intended aim of the text and the background of the author should also be discussed. By
doing so, it is hopefully possible to determine whether the author had any reason to tell
anything other than the truth. The truth being what the author thought to be the truth. It is
however the question whether the information that the author received was correct. It is
therefore also important to establish whether it was possibly a first-hand observation or not.
9
Chapter 4. The evidence
4.1 Vessel remains
This overview is restricted to Hjortspring, the Nydam vessels and the larger Kvalsund vessel
(Kvalsund II). This is because these vessels have been referred to the most throughout the
discourse. Additionally, these vessels formed the basis for Brøgger and Shetelig´s
evolutionary theory on the development of the Scandinavian boatbuilding tradition, which has
been connected to the late use of the sail and which will be discussed in chapter five. A
thorough discussion of other vessels and fragments is outside the scope of this thesis, but
some of them will be discussed in lesser detail in chapter five when necessary.
4.1.1 Hjortspring
4.1.1.1 The excavation
Hjortspring was found in a bog at Als in southern Denmark. Spearheads and an 8-meter-long
plank were found shortly after 1880 during the digging of peat. Unfortunately the spearheads
were left to rust and the plank was dried and used as firewood (Rosenberg, 1937, p. 2). The
find was eventually reported to the Nationalmuseum in Copenhagen in 1920. During the years
1921-22 the site was excavated by Gustav Rosenberg which resulted in a publication in 1937.
Once it became clear the vessel was prehistoric it was decided to only focus on the finds in
the area around the vessel during the first year of excavation and on the vessel the following
year. This was also done in order to test methods of transportation and preservation on other
objects prior to any such attempts on the vessel.
4.1.1.2 Preservation & Conservation
The original shape of the vessel was not preserved, it was distorted and fragmented due to the
overlaying weight of the peat (see figure 2). Due to the digging of peat there were clear 3-5
m2 square cuts in the find layers which had been cut through and removed. An area of 3 m2
had been removed from the planks, cuts at both stems had been made and in several other
places (Rosenberg, 1937, pp. 13, 20). Rosenberg (1937) states that no keel or keel plank was
found and concluded that this must have been the 8-meter-long plank that was taken up
shortly after 1880 (p. 21-22). However, later on, Rosenberg (1937) mentions that several
small fragments of planks with cleats were found amidships which likely belong to the keel
(p. 78).
Figure 2. Hjortspring during excavation. After Rieck & Crumlin-Pedersen, 1988, p. 58.
Chapter 4. The evidence: 4.1 Vessel remains 10
4.1.1.3 Construction & Reconstruction
Rosenberg (1937) was at first uncertain whether a reconstruction could be made due to the
deformation of the remains (p. 20). However, it became clear to Rosenberg that the vessel
wasn´t taken apart prior to its deposition but was deposited as a whole. The location of the
fragments therefore corresponded to their approximate position in the vessel. Additionally,
parts of the framing were made of a harder wood type than those of the planks thus retaining
their shape to some degree.
Hjortspring was built of five large lime planks which overlapped each other, the upper plank
on the outside of the lower one. The planks were ca. 2 cm thick and were sown together with
cord (Crumlin-Pedersen, 2003a, p. 26; Rieck & Crumlin-Pedersen, 1988, p. 70). The planks
had integral cleats to which the framing was lashed. The framing and planking was made as
light as possible resulting in a light-weight vessel and a flexible hull (see figure 3).
Figure 3. Cross-section of Hjortspring after Crumlin-Pedersen, 2003, p. 30.
The remains of Hjortspring were assembled and exhibited. Rosenberg (1937) stated that the
reconstruction should be subject to reservations as regards both the shape and dimensions, of
the vessel as a whole and in many of its details, due to the deformation and fragmented state
of the find (p. 72). A reconstruction drawing was nonetheless made by a naval architect,
Frederik Johannessen, in 1936 (see figure 4) (Crumlin-Pedersen, 2003a, p. 35). This was
based on the measurements taken by Rosenberg during the excavation, photos and the
remains themselves after their conservation (Rieck & Crumlin-Pedersen, 1988, p. 68).
Rosenberg (1937) deemed one could without much concern give credence to this
reconstruction (p. 72).
Due to the weakness of the wood, the parts could not be handled prior to conservation. Any
additional measurements of individual parts seem therefore to have been taken afterwards
(Crumlin-Pedersen, 2003a, p. 23). Changes in the shape and dimensions of the wooden pieces
likely influenced this reconstruction (Rieck & Crumlin-Pedersen, 1988, p. 68; Vinner, 2003b,
p. 134).
As regards the keel or keel plank, it is unclear what exactly was found. Rosenberg (1937)
mentions some fragments of cleats which are likely to be from the bottom plank (p. 78). Rieck
and Crumlin-Pedersen (1988) stated that fragments of the keel were found, but do not specify
their shape or dimensions and where these were found (p. 70). In a later publication by
Crumlin-Pedersen (2003a) it is stated that only fragments of the keel at the ends of the vessel
were found (p. 24). However in a note at the end of the sentence, it is written that “parts of the
Chapter 4. The evidence: 4.1 Vessel remains 11
middle section of the bottom plank” were identified in 1988 (Crumlin-Pedersen, 2003a, p.
24). He thereby confusingly contradicts himself and it is unclear whether Crumlin-Pedersen
refers to the same fragments as mentioned by Rosenberg. Rosenberg (1937) does not
conclude whether Hjortspring had a keel or keel plank. Yet the reconstruction drawing by
Johannessen draws Hjortspring as having a keel plank. Crumlin-Pedersen (2003a) states,
regarding this reconstruction, that for the shape of the keel plank amidships and the stem
projections there is no concrete evidence, but that the margin of uncertainty is small (p. 36).
It is questionable whether it can be concluded on such a basis that Hjortspring indeed had a
flat keel-plank. It is impossible to assess whether Rieck and Crumlin-Pedersen´s conclusion is
correct due to the somewhat confusing and unclear statements. Since no clear information is
given what exactly was found, we are forced to rely on Crumlin-Pedersen´s statement for
now. Whether the conclusion that Hjortspring had a keel plank is based on the actual remains
or on the assumption that the plank retrieved in 1880 was the keel is thus unclear.
Figure 4. Reconstruction and lines plan of Hjortspring by Johannessen. After Crumlin-Pedersen,
2003, p. 36.
Hjortspring underwent a re-conservation and re-mounting during the period 1966-88 and a
full-scale reconstruction was made of Hjortspring in 1971. It became clear that the
longitudinal profile drawn by Johannessen was not entirely correct (Crumlin-Pedersen, 2003b,
p. 209). Another full-scale reconstruction, named Tilia, was made in the 90s which took these
corrections into account giving it a greater rocker (Fenger, 2003, p. 91; Vinner, 2003b, p.
135).
Hjortspring is estimated to have been 18.6-19.6 meters long (stem extensions included), 2,04
meter wide and 0,75 meter high amidships. The interior length would have been 13,61 meters
according to Johannessen´s drawing (Crumlin-Pedersen, 2003a, p. 36).
On top of the stems integral cleats were present which have in the latest reconstruction been
interpreted as attachment points for a trussing rope (see figure 5).
Chapter 4. The evidence: 4.1 Vessel remains 12
Figure 5. The trussing rope fastened to the cleats at the stems. After Hocker, 2003, p. 85.
4.1.1.4 Capabilities
The cross-section amidships is quite circular and therefore lacks initial stability. However,
considerably more force is necessary to capsize it than to heel it 10 (Fenger & Hocker, 2003,
p. 94). Because Tilia was constructed with a greater rocker its ends were less submerged. For
the test trials in 1999-2001 it was necessary to use a ballast of 600 kg in order to increase the
waterline, improving the directional stability and making it steadier (Vinner, 2003a, p. 105).
This was primarily done to improve the manoeuvrability of the vessel because it was found
difficult to steer with only one steering oar at the stern. It proved very manoeuvrable once a
second steering oar was used at the other end (ibid., p. 106). The sea trial in 2000 has also
demonstrated that the reconstruction had a greater cargo-carrying capability and was more
seaworthy than initially assumed being capable of operating in waves of 1 metre high (ibid.,
p. 113). The last sea trial, which covered a longer stretch, has shown that it was possible to
paddle the boat over long distances at an even speed, covering over 40 nautical miles (74 km)
per day in fair weather (ibid., p. 117). The trussing rope was found to be very useful for
working against the trussing of the vessel although it was somewhat of a hindrance for the
crew when moving from side to side.
4.1.1.5 Propulsion
Together with Hjortspring paddles were found of which two were of a larger size. These were
found at the stern and bow. It is now clear that both were needed for steering efficiently,
instead of one being a spare as has been proposed by some (McGrail, 1987, pp. 206, 241;
Rieck & Crumlin-Pedersen, 1988, p. 72). Hjortspring has been interpreted as being solely
propelled with paddles (Rieck & Crumlin-Pedersen, 1988, p. 72; Rosenberg, 1937, p. 89).
Bengtsson (2017) has however recently demonstrated that it is possible to sail with Tilia. He
also argues that the cleats to which the trussing rope was attached would also have been ideal
for the fastening of stays. Hjortspring was capable of sailing without difficulty at 90 degrees
to the wind.
Chapter 4. The evidence: 4.1 Vessel remains 13
4.1.1.6 Date & Provenance
The exact date of the vessel was at first uncertain and different dates were proposed, but all of
these remained within the Pre-Roman Iron Age (500 BC – 1 AD) (Rieck & Crumlin-
Pedersen, 1988, p. 73). In 1987 some minor excavations were done on the site in an attempt to
find wood suitable for 14C dating. The timber found during the original excavation had long
been conserved and thus no longer suitable for a reliable dating. Two fragments of a spear and
a fragment of a frame were found dating to ca. 390 and 370 BC respectively. On the basis of
these, Hjortspring is estimated to have been built in the fourth century BC (Kaul, 2003b, pp.
173–175; Rieck & Crumlin-Pedersen, 1988, pp. 73–74).
The provenance is harder to determine due to a scarcity of finds to compare Hjortspring`s
assemblage to. The objects do seem to show a Scandinavian or Baltic origin (Kaul, 2003b, p.
176). Rosenberg thought it most likely that the crew originated from the western Baltic basin,
possibly the base of the Jutland peninsula. Randsborg (1995) has proposed the Hamburg
region as a likely point of origin on the basis that the wooden containers found with
Hjortspring, share a similar vessel shape to earthenware from the Hamburg region (pp. 65-
66). However, Kaul (2003b) warns that these wooden vessels may have been more
widespread than currently is visible in the archaeological material due to their organic nature
(p. 177). He specifically points at Zealand as lacking finds of ceramics from the first half of
the Pre-Roman Iron Age (500-250 BC) possibly indicating a greater use of wooden vessels.
Another possibility proposed by Martens (2001), based on the find of a shield boss and other
weapon finds in Sweden, is Skåne or Bornholm. It is clear that a precise location of the crews´
origin is uncertain due the scarcity of comparable finds which may also give a distorted view
of their geographical distribution. The general consent seems to be the southern
Scandinavia/northern Germany/western Baltic area, this is partly based on the possible
seafaring range of Hjortspring.
4.1.1.7 Find context
The vessel was deposited in the bog together with a large deposit of weapons. These belonged
to military troops consisting of more people than could be transported with just one
Hjortspring-type vessel (Kaul, 2003b, pp. 178–179). The remains seem to belong to a single
deposition except for possibly a few animals and objects (Kaul, 2003b, p. 142). It has been
interpreted as an offering, as thanks for a victory (Kaul, 2003b, p. 142; Rosenberg, 1937, p.
101). The objects deposited are thought to be of a people foreign to Als who were defeated by
the inhabitants of the island (Kaul, 2003b, p. 142).
4.1.1.8 Function of the ship
Since the vessel was found together with the weaponry of warriors it is believed to have
functioned as a warship or troop transporter (Crumlin-Pedersen, 1990, p. 98; Kaul, 2003a, p.
201; Rosenberg, 1937, p. 91). From sea trials with Tilia it was demonstrated that it could cope
well with 1 metre high waves, being more seaworthy than initially assumed (Vinner, 2003a, p.
113). The symmetrical shape, manoeuvrability, shallow draught and lightness of the boat
enabled easy beaching and disembarking of the vessel (Kaul, 2003a, p. 201; Vinner, 2003a, p.
118). Because of this, its speed (due to its lightness) and its daily radius it would have been
well suited for surprise attacks (Kaul, 2003a, p. 201).
Chapter 4. The evidence: 4.1 Vessel remains 14
4.1.2 The Nydam vessels
4.1.2.1 The excavation
Several vessels were found in the Nydam bog, located on the Sundeved peninsula ca. 10 km
south-west of the Hjortspring bog. Several objects were found in the 1830´s by the owner but
their importance was not yet understood. Another find in the bog was reported in 1858 to the
antiquity collection (Oldsagssamling) in Flensborg which was controlled by Conrad
Engelhardt, the person who led the excavation. The excavations started in 1859 and had to be
concluded in 1863 due to the threat of war, the Second Schleswig War, which started in 1864.
The first traces of a vessel were found on the 7th of August 1863 (Engelhardt, 1865, p. 6). On
the 18th of August 1863 the oak vessel was found and not long after the pine vessel (Rieck &
Crumlin-Pedersen, 1988, pp. 103–105).
The remains of 3 vessels were found, although Rieck and Crumlin-Pedersen (1988) have
argued that remains of a fourth were found (p. 110). Due to the start of the war there was only
time to record and reassemble the oak vessel, the pine vessel had to remain at the bog.
Engelhardt had to write his publication from 1865 without access to the finds (Rieck &
Crumlin-Pedersen, 1988, p. 107).
The vessels are commonly referred to as Nydam A, B, C and D boats. Nydam A being an oak
vessel of which only damaged parts were found, Nydam B the well-known oak vessel, Nydam
C the pine vessel and Nydam D the possible fourth vessel.
4.1.2.2 Preservation & Conservation
The vessel that was best preserved and which is normally referred to in literature as ´the
Nydam boat´ is Nydam B. The keel plank and the five strakes on both sides were preserved,
although the upper strakes on starboard were heavily disintegrated. Large holes were present
in some of the planks under the waterline, possibly made prior to the deposition in order to let
the vessel sink (Engelhardt, 1865, pp. 6, 10). Of the 19 frames that would have originally
been in place, parts of 9 survived. Some are quite complete, but most are fragmented
(Åkerlund, 1963, p. 35). Only one of the thwarts survived. Eight rowlocks and parts of several
more were found (Shetelig & Johannessen, 1930, p. 8).
Figure 6. The preserved parts of Nydam B. After Åkerlund, 1963, p. 51
Chapter 4. The evidence: 4.1 Vessel remains 15
Nydam C (the pine vessel) was found shortly after, lying next to the oak boat. Its remains
were retrieved from the bog, put on land and covered with a layer of soil, with the intention of
transporting it to Flensborg and assembling it as was done with Nydam B. However, the start
of the war prevented this.
It is unclear what exactly remained of Nydam C. Engelhardt (1865) describes it as somewhat
or relatively complete (“nogenledes fuldstændig”) (p. 10). Although being pressured by the
coming war, Engelhardt managed to record its keel, a piece of planking, some rowlocks and
several other objects (see figure 7). The remains, left under the layer of soil, were never
retrieved afterwards (Engelhardt, 1865, pp. 10–11).
Figure 7. Drawing made by Engelhardt of the keel of Nydam C. After (Engelhardt, 1865, p. 11).
The first remains that were found were the cut-up parts from Nydam A. It was observed to be
somewhat different from Nydam B in its construction from the few remains that remained of it
(Rieck & Crumlin-Pedersen, 1988, p. 110). A part of its upper strake and part of another
strake were found during the excavation by Engelhardt (see figure 8). During a later
excavation in 1939 by J. Brøndsted and C. F. Becker a part of its gunwale was found (ibid., p.
113).
Figure 8. The planks recorded from the Nydam A boat. After Engelhardt 1865, figure IV.
It is unclear what type of methodology was applied to preserve Nydam B (ibid., p. 116).
Nevertheless, it is likely that the planks shrunk considerably prior to the assembling and its
recording (Åkerlund, 1963, p. 33).
Chapter 4. The evidence: 4.1 Vessel remains 16
4.1.2.3 Construction & Reconstruction
Nydam A seems to follow the same general construction as the others, i.e. clinker-built with
cleats for lashing the frames to (see figure above). It was presumably smaller than Nydam B
and C (Rieck & Crumlin-Pedersen, 1988, p. 120).
Nydam B was moved to Flensborg and assembled by conservator Stephensen from
Copenhagen in 1863. The boat was recorded by Magnus Peterson. It was clinker-built and
consisted of a keel plank and 5 planks on each side (Engelhardt, 1865, p. 7). The keel and
stems were connected by a horizontal scarf fastened with two treenails (see figure below).
Figure 9. Drawing of Nydam B stem scarf. After Shetelig & Johannessen, 1930, p. 12, redrawn by
author.
It had thwarts which were supported by stanchions. Cleats were present for lashing the frames
to the strakes. The upper cleats had three holes instead of one, two of these were used to
fasten the thwarts with treenails (Åkerlund, 1963, p. 27; Shetelig & Johannessen, 1930, p. 14).
The length of the thwarts was unknown and could therefore not be used to determine the exact
width of the vessel. Of the frames that could have been placed amidships, none were
sufficiently preserved to determine the exact shape of the vessel (Åkerlund, 1963, p. 37). Prior
to the assemblage the process of drying and thereby the shrinking and warping of the remains
had already started. This has mainly influenced the initial recording of the width, i.e. being
less than it originally would have been. Another reconstruction was made in the 1930´s by
Johannessen, which had become the accepted reconstruction drawing. This was more a
measurement of the state of Nydam B boat in the 30´s rather than a reconstruction. It becomes
clear that the shrinking of the timbers was not taken into account when comparing the
drawings made by Peterson and Johannessen (see figure 10). This had however gone
unnoticed until it was pointed out by Åkerlund (1963). Åkerlund (1963) estimated that the
planks had shrunk 13-14% in width (p. 35). He proposed a new reconstruction drawing which
takes into account the shrinking and the uncertainties during the initial reconstruction in
Flensborg. This resulted in a wider and rounder cross-section, a rocker-keel and rounded
stems.
Chapter 4. The evidence: 4.1 Vessel remains 17
Figure 10. The different reconstructions made of the cross-section amidships. After Åkerlund, 1963, p.
36.
Åkerlund mentions several mistakes made during the first or second reconstruction. The holes
in the cleats and frames through which the lashing went were misaligned, the gunwales were
fastened to the upper strakes through different holes than it original would have and some of
the frames were not placed in their correct position (Åkerlund, 1963, p. 36). It is unsure when
this happened. Shetelig and Johannessen, in their article of 1930, mention that some mistakes
were made during the first assemblage and that these were corrected by them during its
assemblage in Kiel (p. 6). It is however unclear what kind of mistakes they refer to. Åkerlund
(1963) points out that the frames were no longer in their original position when excavated and
could thus already have been misplaced during the initial assemblage (p. 37).
Since Nydam B boat has a keel plank it is by many considered to have lacked sufficient
longitudinal strength for rougher waters (Shetelig & Johannessen, 1930, pp. 20–21). Its v-
shaped cross-section was thought to have been a solution to avoid too much pressure on the
keel. Since Åkerlund has proposed a rounder cross-section, thus having more pressure on the
keel, he proposes that a construction would have been in place in rougher waters to counteract
the sagging and hogging forces on the vessel. Åkerlund is of the opinion that certain wooden
objects found in both Nydam B and C, of which no clear function has been put forward, were
used for this construction. Engelhardt (1865) mentions that many rounded staves were found
in between the frames of the vessels and also some wooden objects with a hole in each end (p.
6). Åkerlund (1963) proposed that the staves were inserted through the holes to help against
the sagging and that ropes running along the upper part of the construction were used to
counteract the hogging of the vessels (see figure 11) (pp. 63-73). Åkerlund is of the opinion
that Nydam B had through-beams at its ends which were used to attach this longitudinal
construction to.
Chapter 4. The evidence: 4.1 Vessel remains 18
Figure 11. The longitudinal construction as imagined by Åkerlund. After Åkerlund, 1963, figure IV.
Rieck & Crumlin-Pedersen (1988, p. 118) mention that several objections can be made to
such a construction, although they only specifically mention one. Namely, that such a
construction is not visible in the contemporary iconography of the period, likely referring to
the early Gotlandic picture stones. Another objection by them is that it is unpractical, but this
is not explained any further. As mentioned for Hjortspring, the trussing rope was somewhat in
the way, but it was clearly preferred by the crew and seen as essential in a later sea trial (Boel
Bengtsson, 2017, p. 106). Unpractical side-effects may be deemed to outweigh the positive
effects. Nonetheless, this construction remains uncertain.
Rieck & Crumlin-Pedersen (1988) seem to accept the new cross-section as proposed by
Åkerlund, but not the longitudinal shape (p. 118). The sideview as drawn by Åkerlund has
more of a curve than the previous line plans because he believed the keel had a rocker. Since
the planks of the vessel had no scarfs (except for the upper strakes) and were of a large
dimension, Rieck & Crumlin-Pedersen argue that only straight grown trees could be used.
They argue therefore for a longitudinal shape that is produced with only straight planks, i.e.
less curved (ibid., pp. 118–119).
According to Engelhardt (1865) Nydam C (the pine vessel) was built like Nydam B (p. 12). It
was clinker-built, fastened with nails and roves and had cleats for attaching the frames to. Not
much is known of the vessel except from the drawings and short description by Engelhardt.
Åkerlund has also made a reconstruction of the pine boat based on the drawings of its keel
and a drawing of part of a plank. Åkerlund defines the 15,65 meter long keel plank of the pine
vessel as a keel, since it has a distinctive keel cross-section (see figure below) (1963, p. 94).
Figure 12. Keel section of the Nydam pine boat. After Engelhardt 1865, p. 11.
Chapter 4. The evidence: 4.1 Vessel remains 19
Midships the keel was 34 cm wide and 12 cm thick with a T-shape and is almost a fully
developed T-keel (Åkerlund, 1963, p. 98). A similar type of keel is also present in the later
Kvalsund vessels. Nydam C is the oldest known Scandinavian vessel with a “real” keel
according to Åkerlund (1963, p. 98). The scarf between the keel and the stems was also
different from Nydam B. The ends of the keel run to a point. Engelhardt couldn´t comprehend
how this would fit in the vessel´s construction and proposed that the pointed end was covered
with sheets of metal and used as a ramming device. Åkerlund (1963) has however put forward
a more logical explanation based on a somewhat similar ending on the Björkebåt (a vessel
from the 5th century) (p. 95). It is now thought that it likely had a stepped stem (see figure 13).
Figure 13. How Åkerlund imagined the ends were constructed. After Åkerlund, 1963, p. 99.
It is unknown how many strakes Nydam C would have had, but since they are less wide,
probably more than Nydam B. Åkerlund (1963) estimates seven strakes on each side (p. 98).
4.1.2.4 Capabilities
Prior to Åkerlund´s observation, the recording of Johannessen from 1929 was used in the
discussion regarding the capabilities of the boat. Due to the shrinking, the width in
Johannessen´s drawing was 4 metres instead of the 4,4 metres as recorded by Engelhardt. This
gave Nydam B a sharper cross-section, being seemingly less stable than it in reality was.
Shetelig and Johannessen believed it would have been very unstable. Stones were found in
Nydam B which Shetelig and Johannessen (1930) therefore interpreted as ballast to increase
the stability (p. 22). Åkerlund is of the opinion that this would not have been necessary based
on his reconstructed lines plan and that these stones were simply used to sink the boat.
Åkerlund (1963) points out that it is unclear how many stones were found (p. 19).
Johannessen determined that Nydam B´s weakest point was its keel. It was therefore thought
that the boat had steep sides on purpose to decrease the forces on its keel (Shetelig &
Johannessen, 1930, pp. 21–22). The notion that de steep sides were necessary to compensate
for the weak keel had also been widely accepted and repeated by others (Åkerlund, 1963, p.
46). Since Nydam B had a rounder cross-section Åkerlund acknowledges that its keel could
not cope with the forces of the sea. However, the construction as proposed by Åkerlund, for
both the oak and pine ship, would help against these forces. On the other hand, it is unsure
whether such a construction was indeed in place. The main argument of Åkerlund is that it
would have needed it in rougher seas. Tuxen, another expert, based his view on the drawing
of Engelhardt and stated that the Nydam oak boat was well suited for coastal waters, such as
between the Danish islands, or in the Norwegian archipelagos and fjords (ibid., p. 45). Nydam
Chapter 4. The evidence: 4.1 Vessel remains 20
B is clearly seen to have been restricted to coastal waters and not suited for oceanic voyages
(Jones, 1996, p. 81).
Åkerlund´s reconstructed cross-section of Nydam C is hardly reliable. The keel of Nydam C
would however give better longitudinal strength than that of Nydam B. Additionally,
Åkerlund (1963) states that the scarf between the keel and the stems of Nydam C is much
stronger than that of Nydam B which in his opinion is quite weak (p. 22, 95). However,
Shetelig and Johannessen (1930) are of the opinion that the scarf of Nydam B is strong (p. 9).
It should also be said that the scarf of Nydam C is only a proposed solution by Åkerlund.
4.1.2.5 Propulsion
Engelhardt (1865) concludes that the vessels were likely rowing boats and did not carry a sail
(p. 12). The reason for this is because no masts and no facilities were found in the vessels to
attach ropes to. This has also been repeated by others afterwards (Åkerlund, 1963, p. 22;
Brøndsted, 1940, p. 236; Dahlgren, 1932, p. 75; Shetelig & Johannessen, 1930, p. 15).
There was a hole in both Nydam B and C in the middle of their keels midships. These had a
diameter of 1.5 Danish thumbs (3,92 cm) and went all the way through the keel (Fig. 12). It
would therefore more likely have functioned for the draining of water rather than a mast-step.
The conclusion that the Nydam vessels were propelled solely by oars seem to be accepted by
all.
Nydam B was propelled by 28 or 30 oars (Åkerlund, 1963, p. 77). The oars are shorter than
those from later periods which means that only smaller strokes were possible. The distance
between the thwarts and the oarlocks would be 25-28 cm if a shift in bow and stern as
proposed by Åkerlund is accepted. This would be too short for a full stroke with today´s oars,
but would fit the shorter oars (Åkerlund, 1963, p. 56). Side rudders have been found for both
Nydam B and C (Crumlin-Pedersen, 2010, p. 67; Engelhardt, 1865, p. 8).
4.1.2.6 Date & Provenance
It is unsure whether the vessels are contemporary (Rieck & Crumlin-Pedersen, 1988, p. 120).
Engelhardt thought so, but Åkerlund (1963) is of the opinion that they were separate
depositions because of their structural differences (pp. 104-116). Rieck and Crumlin-Pedersen
believe that the vessels originated from three different locations each being a variant of the
northern Germanic boatbuilding tradition. They argue that, because large trees were required
for the construction, the vessels were built in the core areas of their corresponding tree
species. Thus Nydam C, constructed of pine, could not have been built in Denmark, but rather
in Western Norway, Mid-Sweden or the southern Baltic coast (Rieck & Crumlin-Pedersen,
1988, p. 120). Based on the same thought, the Nydam A and B would have been built in
Southern Scandinavia or at the southern Baltic coast. Rieck and Crumlin-Pedersen look
further at the constructional differences to determine the origin of the vessels. According to
them the archaeological material indicates that the rabbeted stem is only used at the North Sea
whilst the stepped stem is mainly found in the Danish and West Swedish Viking period ships.
They therefore think it more likely that Nydam B is from the area around the North Sea. This
is based on vessels from different periods and it is questionable how reliable this is. Rieck and
Crumlin-Pedersen (1988) acknowledge that the amount of ships on which this theory is based
is too small (p. 122). Nonetheless, they conclude that the Nydam vessels indicate either three
separate depositions or a geographically mixed army.
Åkerlund (1963) is of the opinion that the difference in construction of Nydam B and C points
to a large time difference, Nydam C being the youngest as it has the most advanced
Chapter 4. The evidence: 4.1 Vessel remains 21
construction (p. 105). Åkerlund estimates that this might have been a few hundred years but
states that he does not exclude the possibility that they are from different regions. It is
however clear that he is convinced that they are from different periods. Åkerlund (1963)
namely concludes that “it is seems hardly justifiable, without more compelling reasons, to put
it the time difference at less than a hundred years” (p. 105, own translation). The dating of
the Nydam vessels was based on the typology of the fibulae giving a date somewhere around
300 AD. Åkerlund (1963) therefore concludes that:
This would give us the right to assume that the oak boat may be lowered in the bog
sometime during the start of the 4th century and if we accept it as likely that the pine
boat was sacrificed at the beginning of the 5th century then we keep the acceptable age
difference between these boats with regards to their technical quite different
constructions. (p. 116, own translation)
There are several issues with the above mentioned reasonings. They are based on several
assumptions that should only be accepted after being underpinned with arguments. The
assumption that vessels with different constructional features belong to different period as
stated by Åkerlund is clearly based on an evolutionistic view which excludes any possibility
of regional diversity. It may thus be an army comprised of warriors from several regions as
Rieck and Crumlin-Pedersen propose. The availability of resources as they argue, may indeed
point to the region were the vessels were constructed. On the other hand, this does not
necessarily imply that the army consisted of a geographically mixed army. The same group of
people may also use several different types of vessels, which were acquired by other means.
Although both Nydam A and B are both from oak, Rieck and Crumlin-Pedersen do not seem
to think it possible that they are from the same group due to the difference in construction.
The idea that one group can only employ one method of construction at the same time is a
recurring thought in the literature which is simply an assumption with no clear basis.
4.1.2.7 Find context
Together with the vessels, weapons, equipment and personal belongings were deposited. It is
thought that these objects belonged to an attacking army that was defeated by the local
inhabitants as was the case with Hjortspring (Åkerlund, 1963, p. 122; Shetelig &
Johannessen, 1929, p. 54) . Some of the weapons were destroyed as was Nydam A. Since
Engelhardt (1865) thought that the vessels belonged to the same deposition he was of the
opinion that lack of time was the reason why the other vessels were not destroyed, but simply
sunk by cutting holes in the sides of the vessels (p. 5). Shetelig thought that such holes would
have been made in other places and larger if this was indeed the intention (Shetelig &
Johannessen, 1930, p. 4). In addition to the Nydam offering other depositions were made in
the bog throughout the 3rd and 4th centuries (Rieck & Crumlin-Pedersen, 1988, p. 115). It is
thought that these kinds of depositions in holy lakes or bogs were made as an offer to the gods
as thanks for a victory (Åkerlund, 1963, p. 18; Rieck & Crumlin-Pedersen, 1988, p. 102).
4.1.2.7 Function of the ship
The Nydam vessels have been interpreted, based on their find context, as warships or troop
transporters as it was the aim to build a ship that could carry as many people and oars as
possible (Crumlin-Pedersen, 1990, p. 98; Shetelig & Johannessen, 1930, p. 30; Vogel, 1915,
p. 63). It is also thought they had to be capable of withstanding the forces of beaching on
sand, ergo the thickened middle part of Nydam B (Greenhill, 1976, p. 181). As mentioned it is
thought to be a pure rowing vessel and as such was narrow, had little draft and did not have a
“proper” keel (Christensen, 1996, p. 77).
Chapter 4. The evidence: 4.1 Vessel remains 22
4.1.3 The Kvalsund vessels
4.1.3.1 The excavation
On the first of June 1920 in Kvalsund in the Møre og Romsdal province of Norway, timbers
were found during the digging of peat. The farmer recognized it as parts of a vessel and the
find was reported to the museum in Bergen. On July 2nd, Shetelig came to Kvalsund to
excavate the site. Several fragments were already exposed by the digging. These showed a
haphazard distribution with stones on top and under these fragments (Shetelig & Johannessen,
1929, p. 13). Two vessels were deposited, a smaller and a larger one, from here on referred to
as Kvalsund I and Kvalsund II respectively.
4.1.3.2 Preservation & Conservation
During the excavation the wood was protected by wet peat and covered with wet bags. After
the fragments were exposed and cleaned they were taken up and immediately treated with
carbolineum and linseed oil (Shetelig & Johannessen, 1929, pp. 15–16). Shetelig remarks that
the preservation was not ideal, because the finds had not been lying deep in the bog which
was neither very wet. The wood was quite rotten and fragile and of a porous and weak
consistency. The wood was slowly permeated by the carbolineum and linseed oil as the water
was removed, thus strengthening the consistency whilst the volume and shape remained
unchanged according to Shetelig and Johannessen. After a long drying process wax or lacker
would be applied (ibid., p. 16).
The clear majority of the fragments are the remains of Kvalsund I (the smaller vessel). Every
single part of the vessels was broken into smaller pieces, of which some were damaged by fire
prior to their deposition. The distribution of these fragments looked at first quite random, but
there seems to have been some structure with the long pieces laid side by side and the shorter
thicker pieces laid across (ibid., p. 29).
4.1.3.3 Construction & Reconstruction
Johannessen was contacted by Shetelig to record and examine the vessels more closely and to
establish whether a reconstruction drawing could be produced. Although a find list with short
descriptions is present in the publication, it is quite unclear how much was preserved of each
vessel. This is due to descriptions lacking information and because it is unclear to which of
the vessel each fragment belongs. Johannessen does make clear on which fragments his
reconstruction is based. However, it is difficult to verify his reconstructions and one is
somewhat forced to believe his conclusions.
Figure 14. Reconstruction drawing of Kvalsund II by Johannessen. After Shetelig & Johannessen,
1929.
Chapter 4. The evidence: 4.1 Vessel remains 23
Kvalsund II was more difficult to reconstruct than Kvalsund I because less was preserved of
it. Nonetheless, Johannessen mentions that enough was present to produce a reconstruction
and gives an explanation of how the reconstruction came about and why it is the only possible
solution (Shetelig & Johannessen, 1929, p. 60).
Johannessen shortly describes the remains that were of importance to his reconstruction;
Of the keel several parts remained both from amidships and its ends, of which the
foremost had a scarf for the stem; several frames in larger and smaller fragments, the
rudder frame´s complete starboard side and the rudder; partly large lengths of strakes
with cleats for lashing, and not the least notable: two brander. They were indeed
broken pieces, but could easily be matched to the fractures, so at least one of them
could be accurately drawn in its correct form. (ibid., pp. 60–61, own translation)
Brander are the uppermost decorative strakes at the stems (see figure above).
Most of the frames found were from the midship section which gave the ship´s width and
depth. The length could not be established based on the fragments. For an estimation
Johannessen looked at the length-breadth ratio of other known ships and takes the smallest
and the largest ratio as the outer limits. The smallest being Gokstad with a 4,66 ratio and the
largest being Kvalsund I with a 6,37 ratio. He states that there are no reasons to exceed these
limits. This gives a length anywhere between the 12,6 and 20,4 metres. Based on the midship
frame Johannessen states, although it is unclear why, that it would be nearer the L:B ratio of
Nydam B and Kvalsund I (ibid., p. 61). Other frames belonging to the midship section are
quite similar in length as the supposed midship frame, Kvalsund II would therefore have had a
rather straight midship section. The cleats on the planks indicated the distance between the
frames. These details in combination with the dimensions of the material, especially the keel,
the placement of the brander and a reasonable trajectory of the gunwale line have given, as
the reconstruction drawings came along, the vessel´s length and shape (ibid., p. 62). This
resulted in an estimated length of 18 metres, a width of 3,2 m and a depth, measured to the top
of the keel, of 0,785. This gives a L:B ratio of 5.62.
No stems belonging to Kvalsund II were found during the excavation. Their shape in the
reconstruction drawing has been determined by making sure that the brander was placed in
such a position that the rabbet of the keel going towards the end of the brander and the line of
the gunwale going to the start of the brander formed smooth curves. It is however clear that
Johannessen´s idea of a correct line was based on his knowledge from previous
reconstructions. He states that it was the aim that the lines formed the smooth curve, known
from the Oseberg ship and from carvings and drawings (ibid., p. 63). It should therefore be
questioned whether the reconstruction was based on the actual fragments or was biased
because of Johannessen´s previous reconstructions of the Viking ships. It is difficult to judge
how accurate his reconstruction is since no drawings of individual parts were published and
we are therefore forced to trust his words as is the case with many of such early
reconstructions. It is also unclear what exactly remained of the vessel since only those parts
deemed important for Johannessens´ reconstruction are mentioned. The description of the
vessel is rather a presentation of his conclusions rather than a description of the remains.
The keel is T-shaped, the widest part being 20 cm tapering down towards the stem scarfs. The
scarf is similar to Nydam B with the stem scarfed on top of the keel. The keel has cleats for
Chapter 4. The evidence: 4.1 Vessel remains 24
each frame but these have no holes for attaching them. The frames are made from one piece,
running from one gunwale to the other. They are made of pine and are according to
Johannessen large in size. He thinks that this was possibly so because only loose thwarts were
used. No evidence of them having been fastened was at least found on the frames or on the
thwart ends that were found. It is difficult to verify this statement since it is unknown how
many frames and thwart ends were found and where exactly these were located in the vessel.
It is stated that only a few fragments were preserved of the thwarts. No beams were found.
Johannessen mentions that if the ship would have used a sail it would have needed a
transverse support at the mast which then seems to have been the only one (ibid., p. 63).
Whether the absence of beams from the find indeed indicates that it did not have any, as many
have stated afterwards, is questionable. As mentioned it is unclear how many fragments were
found of Kvalsund II. Since neither the stems were found one should be cautious with such
conclusions. A beam can simply be fastened with treenails to the frames leaving few traces.
According to Johannessen, Kvalsund II had eight strakes with cleats on each side. These were
made of oak, except for the upper strakes which were of pine. There are six holes in the
frames on each side for attaching the planks by way of lashing. The 5th and 6th strake are both
tree nailed and sown, the 7th strake is only tree nailed and the upper strake is fastened with
nails that are clenched on the inside.
4.1.3.4 Capabilities
Jones (1996) has stated that experiments with a replica of Kvalsund II “revealed rather
alarming features and suggested their limitations as vessels for crossing hundreds of miles of
open water and weathering North Sea gales.” (p. 81). It is unclear were Jones got his
information from since no reference is given that mentions any experiments with Kvalsund II
and no literature was found mentioning such an experiment.
4.1.3.5 Propulsion
Oars were found belonging to Kvalsund II. The width of the vessel would have meant that one
person per oar was used (Shetelig & Johannessen, 1929, p. 65). It has been constructed with a
total of 20 oars. Regarding the possibility of it having had a sail Johannessen states:
There is nothing present that can inform us whether it used a sail, but its size, the
frame dimensions and the rudder´s large depth – in order to stay in the water whilst
heeling to port – indicate that this could have been the case, whilst its low freeboard
indicates the contrary. … the question whether it had a sail or not must be said to be
insoluble. (ibid., p. 65, own translation)
However throughout the text there is mention of a mast by Shetelig (ibid., pp. 29, 30, 34).
Johannessen does not discuss the possible mast or considers it as a possible indication for the
use of a sail as becomes clear from the quotation above.
The fragments of the possible mast are listed in the find list as no. 26, 85 & 90. No. 26 was a
5-metre-long round piece of pine, heavily eroded, but it was possible to see that its original
surface was smooth and finely shaped. At one end it has a diameter of 6,8 cm and at the other
a diameter of 11,5 cm. Object no. 85, 67 cm long, possible fits to the wider end, but due to
erosion it was not possible to see whether this was the case (ibid., p. 70). Object no. 90 is a 30
cm long round piece of pine. One end is well preserved which has a diameter of 12 cm. It is
stated that it is likely a part of no. 26, since the dimensions fit, but due to erosion it was not
possible to fit them together. The possible mast would then have been 5,75 metre tall (ibid., p.
Chapter 4. The evidence: 4.1 Vessel remains 25
70). Shetelig doesn´t seem to exclude the possibility that Kvalsund II had a mast and sail in
any of his later publications.
4.1.3.6 Date & Provenance
Shetelig dates the Kvalsund vessels to somewhere between the 5th and 8th centuries by
comparing its constructional features to the dated Hjortspring, Nydam and Oseberg vessels
(Shetelig & Johannessen, 1929, pp. 55–56). Later on it was 14C dated to ca. 690 AD
(Christensen, 1984, p. 130). The provenance has not been discussed, but it is seems implied
that it was a type of vessel commonly used in the region where it was found.
4.1.3.7 Find context
According to Jens Holmboe (1929), who did the botanical research, the bog had never been a
lake (pp. 17-21). The Kvalsund vessels were thus deposited in a bog in contrast to Hjortspring
and the Nydam vessels which were deposited in water that became a bog later on. However,
there are some indications that the pit which was dug for the deposition of the vessels, had
stood open for a while (Jessen, 1929, p. 24). In the lower layers (below the boat fragments)
wood shavings were present (Shetelig & Johannessen, 1929, p. 27). Under and partly in the
find layer, rounded stones were present in varying sizes, but none too big to lift. These did not
form a clear layer, but rather a random distribution.
A lot of branches and twigs, which were thrown in and stuck through the peat layer, were
found in the find layer in a random distribution. Most of the branches had sharpened ends on
them. Some of the fragments of the boat were stuck vertically into the ground. The oars were
stuck into the ground in a slanted position similar as when in use on the vessel (ibid., p. 30).
Shetelig and Johannessen proposed that the deposition of branches with their sharpened ends
stuck into the ground are meant as replacements for weapons. Branches stuck into the ground
have also been observed in other bog finds (ibid., p. 39).
Almost all the parts of the vessels had been destroyed prior to the deposition. Only the rudder
and two oars have remained complete. They seem to have avoided the use of tools, the main
method was simply breaking the parts by force (ibid., p. 34). This is clearly a ritual
destruction which is also indicated by its location where several large burial mounds from the
Bronze Age are present, indicating that it was a place of worship (ibid., pp. 34–35). Due to the
deposition of the vessels in close connection to these monuments, Shetelig and Johannessen
(1929) interpret it as an offering (p. 36). Since there are quite some parallels with the Danish
bog finds, it may be that the Kvalsund deposit also represents an offering after a victory (ibid.,
pp. 40, 54).
4.1.3.8 Function of the ship
Due to the find context it has been interpreted by Shetelig as being a vessel used for war
(Shetelig & Johannessen, 1929, p. 54).
Chapter 4. The evidence: 4.2 Iconography 26
4.2 Iconography
4.2.1 The Gotlandic picture stones
4.2.1.1 The stones
Standing stones with representations of vessels on them are present on the Island of Gotland,
to the east of southern Sweden. These were analysed by Sune Lindqvist and published in 1941
and 1942, for the first time giving a comprehensive compilation of the picture stones known
at that moment. Lindqvist ordered them in a typological sequence. On the basis of shape,
style, depictions, decoration along the edge and the technique used, he divided the stones into
several groups; A, B, C, D, and E.
Figure 15. Showing the typology in shape and their respective dating by Lindqvist. After Nylén &
Lamm, 2003, p. 172.
Type A stones are signified by circles decorated with different types of spirals, stylised
animals seemingly representing horses, and curved vessels propelled by oars with steering
oars both fore and aft. The sides of the stones curve slightly outwards at the top. The borders
of the stones are decorated either with simple lines or more intricate motifs. In a few
examples people are shown with weapons or fighting some sort of animal/mythological
figure. These stones have been dated by Lindqvist to the period 400-600 AD (Imer, 2004, p.
48; Lindqvist, 1941, pp. 108–115).
Figure 16. A type A stone with a representation of a ship. The Bro I picture stone. After Lindqvist,
1941.
Chapter 4. The evidence: 4.2 Iconography 27
Type B stones are smaller than the type A stones and have the nickname ´dwarf-stones´.
Furthermore, their sides are more curved, their decoration is more stylised and they depict
different scenes than type A stones. Most of the vessels are now depicted with a mast and sail
and with angular stems. They are several times depicted together with birds and sometimes a
snake or a horned animal (stag, reindeer, elk?). These are dated by Lindqvist to 500-700 AD
(Lindqvist, 1933, p. 107, 1941, pp. 115–117).
Figure 17. Type B picture stones. From left to right: Rikvide, Fole K & Endre Skog. After Lindqvist,
1941.
Type C and D stones are larger than type B stones and have a mushroom-shaped outline.
Vessels are depicted more detailed with rigging and with people onboard. These vessels are
normally placed on the lowest part of the stone. Above the vessel depictions several scenes
can be seen. Group C was dated by Lindqvist to the first half of the 8th century and group D to
the latter half but could also be from the 9th and 10th centuries. The C and D groups were
differentiated based on their edge decoration and because the figures on the type D stones
were carved deeper than those of type C (Imer, 2004, p. 48).
Figure 18. The Lärbo St. Hammars I picture stone to the left and the Hejnum Riddare stone to the
right. After Lindqvist, 1941.
Chapter 4. The evidence: 4.2 Iconography 28
4.2.1.2 Meaning
These stones are thought to have been memorial stones. The type A and B stones were part of
grave monuments, whilst the later type C and D stones were often placed beside roads and in
open spaces in groups (Nylén & Lamm, 1988, p. 9). However, some of the later types can be
associated to burials beneath the ground surface (ibid., p. 13). Most of them are not found in
their original context, but as building stones in later period graves or in churches. The picture
stones of types C and D depict mythological scenes. The vessels have been interpreted as
depicting ships for the dead to travel on (Imer, 2004, p. 48; Nylén & Lamm, 1988, p. 15).
Crumlin-Pedersen (2010) has argued that depictions of vessels mostly represent the ship of
Freyr called Skíðblaðnir, but other mythological ships may also be depicted if other elements
in the scene indicate this (pp. 145-157). It is difficult to assert what meaning was attached to
the depictions on the stones, but it is quite certain, taking into account the context, that they
depicted religious and mythological scenes (Imer, 2004, p. 48; Nylén & Lamm, 1988, pp. 16,
17).
4.2.1.3 Dating
It seems necessary to provide a fairly thorough explanation of how the type B stones are
dated. Different dates are given in the literature and several times these are incorrect because
of either a misunderstanding of previous literature or by referring to publications that mention
an incorrect date (McGrail, 2001, p. 211; Westerdahl, 2015, p. 48). Depictions are sometimes
shown which are said to be the earliest depictions of sailing vessels although they are from the
later C or D type stones (Greenhill, 1976, p. 200; Westerdahl, 2015, pp. 40, 41, 48).
The dating of the picture stones is a difficult task with many uncertainties. Only a few of the
stones are found in a datable context. For the date of the picture stones people continue to
refer to the Lindqvist´s dating and typology. Lindqvist dated the earliest stones to the 5th
century. Some scholars were at first sceptic towards this early dating but the date was
confirmed with a find of an unfinished picture stone used in a grave dating to the 5th century
(Nylén & Lamm, 1988, pp. 158–159). The stone must be contemporary with or older than the
grave.
Although several researchers have called for a re-dating of some of the types (Imer, 2004, p.
47), no real changes have been made until Varenius in 1992. Varenius main focus was on the
vessels. His analysis is based on 17% of the then known picture stones. He divided these
stones into four groups based on the representation of the vessels. These are (after Varenius,
1992, p. 141);
- I. Ships with no rig (5th-6th century)
- II. Ships with a simple sail, sheets in corners allowed (7th-9th centuries)
- III. Ships with sail and sheets in a Y-shape along the bottom-line (9th-10th centuries)
- IV. Ships with sail and sheet net (10th century)
Ships of Varenius´ group I are found on type A stones but a few also on type B stones, group
II mainly on Lindqvist type B stones, group III on type C, D and E stones and group IV only
on type C and D stones (Imer, 2004, p. 93) (see figure below).
Chapter 4. The evidence: 4.2 Iconography 29
Figure 19. The overlap of Varenius and Lindqvist´s divisions. After Varenius, 1992, p. 56.
Figure 20. The divisions made by Varenius and Lindqvist with their respective date. Made by author.
Varenius thus dates the first depictions of the sails, type B in Lindqvist´s typology, to the 7th
century instead of the 6th century. Lindqvist dated the type B stones on the basis of comparing
the decoration on the stones to those found on objects from Gotland. At first he attempted to
date the stones on the basis of the edge decoration which has a typical z-pattern that later on is
replaced by more complex patterns. He found the same z-decoration on some metal objects
dating to 600-750 AD (see figure below).
Figure 21. The bronze brooches referred to by Lindqvist. After Nerman, n.d., fig. 81 & 87.
He states that the pattern becomes more common during the Viking Age and is therefore not
suitable for dating the type B stones. There is another type of edge decoration with an extra
line between the Z/S (see figures below). There are however no other objects with a similar
decoration. He therefore tries to date stones that are decorated in Salin Style II and stones
with similar band decorations. These are; Sanda Sandegård I, Fole K, Endre Skog, Stenkyrka
IX, Lärbro Norder-Ire I and Hellvi St. Ire III.
Lindqvist 1941
Type B
Type A
Type C, D & E
5th c.
6th c.
7th c.
8th c.
9th c.
10th c.
Group I
Group II
Group IV
Group III
Varenius 1992
Chapter 4. The evidence: 4.2 Iconography 30
Figure 22. Gotlandic picture stones in Salin style II and with similar band decoration. From top left to
bottom right: Sanda Sandegårda I, Fole K, Endre Skog, Stenkyrka IX, Lärbo Norder-Ire I and Hellvi
St. Ire III. After Lindqvist, 1941.
Lindqvist points out the similarities with some of the band weaves on these stones to
decoration found on bronze fibulae dated to 550-675 AD (see figures below).
Figure 23. The bronze fibulae referred to by Lindqvist. After Nerman, n.d., fig. 23, 24 & 27.
Lindqvist (1941) refers to some bronzes in Salin´s work (1904, fig. 545, 548, 552, 556) found
on Gotland which have, according to him, a strong likeness with the heads on the
Chapter 4. The evidence: 4.2 Iconography 31
aforementioned stones (p. 117). But it is not possible to find full agreement between the forms
on the stones and those on the bronzes and it is therefore not possible to give an exact date for
Stenkyrka IX. Lindqvist states that there is no firm starting point for the dating of the stones
in type B. However, based on the comparisons made he concludes that type B covers at least
the greater part of the 6th and 7th centuries (Lindqvist, 1941, p. 117).
Varenius is sceptical towards the usage of decoration styles for dating. He therefore only
wants to rely on stones dated by context or runes. This is however problematic since only a
few stones were found in a dated context. Two stones were found in a context giving a
terminus ante quem dating. One of these is a group II stone found in a grave dating to the 10th
century. This says however little about the starting date of these types of depictions. Another
find is the Buttle Änge stones which are relevant for the chronology of the type C stones. Imer
(2004) states that Varenius´ dating of group II to the 7th century is based on these stones (p.
93). This is however not the case since there is no mention by Varenius of either Varenius
group II or Lindqvist type B, only of Lindqvist type C (Varenius, 1992, p. 81). Varenius
doesn´t seem to give any explanation for proposing a later date for the stones with the first
representations of vessels with sails than Lindqvist. As mentioned, Varenius also dates the
stones based on runes. He mentions that 7% of the group II stones has runes but does not
mention these in the section concerning the dating (ibid., p. 62).
A possible reason for proposing a later date, may be that Varenius is of the opinion that the
type B stones depicting vessels without sails, which he classifies as group I, are of an earlier
date than type B stones depicting vessels with sails. This seems to be indicated in Varenius´
comparison of his groups to the types of Lindqvist (see figure 18). He provides however no
argument to support this. He does mention that he is of the opinion that the stones showing
ships without rigging represent a category on its own, seemingly implying that this also
reflects a chronological difference (ibid., p. 59).
There are four type B stones which Varenius believes do not have a sail represented. These
are; Endre Skog, Sjonhem Lilla Sojvide, Roma kloster and Halla Broa VII A. This can be
questioned since Endre Skog has a decoration right above the vessel which can be interpreted
as a stylistic representation of a sail (see figure 22). Varenius (1992) rejects this possibility
since it has no square surrounding it (p. 58). Halla Broa VII has a vessel with no sail present
on one side, but on the other side it does have a vessel with, what seems to be, the beginning
of a mast. However Varenius (1992) excludes the representations with only a mast from his
analysis thereby excluding possible evidence that speaks against the vessels with no rigging
belonging to one category (p. 56).
It should be questioned whether the ´boundary´ should be moved by as much as 100 years
later based on these four stones. Furthermore, there are no type B stones dated by context.
Varenius thus still relies on the dating set up by Lindqvist.
It is clear that the later dating to 600 AD as given by Varenius cannot simply be accepted as
accurate. On the other hand neither can Lindqvist´s date. It is simply very uncertain. The only
indications of a date are provided by the Sande Sandegård I and the Stenkyrka IX stones.
These clearly belong to Salin style II which is dated to ca. 560/70-800 AD (Solberg, 2003, pp.
184, 194–195). The bronze objects from Gotland (fig. 23) with similar decoration as the
stones in figure 22 are dated to 550-675 AD. Although the edge decoration cannot be used to
date the stones they do seem to develop from simple to more complex (Imer, 2004, p. 100). If
Chapter 4. The evidence: 4.2 Iconography 32
this is the case then the other stones with the simpler z-pattern are younger than the stones in
figure 22.
The possibility of the first sailing vessels dating to 600 AD as proposed by Varenius cannot
be disregarded, but neither can the date of 500 AD since Varenius provides no clear argument
against this. It is based on the theory that the figures with and without sail cannot have been
produced at the same time, i.e. that the change shows a chronological change. Varenius
divides and orders the Gotlandic picture stones on the basis of this theory prior to the analysis.
As mentioned he provides no evidence for the date of his group II, but he does date it. The
evidence is thus rather fitted to the theory rather than the theory to the evidence. Additionally,
the Halla Broa VII stone with a vessel with no rig on one side and a vessel with a possible
mast on the other may be used as evidence against his theory.
4.2.2 The Häggeby stone
In mainland Sweden there is a stone with a similar rowing vessel to that of the type A stones
in Gotland. On the other side one sees two horses facing each other with a person behind each
of them. Similar scenes of horses facing each other are also visible on the Gotlandic pictures
stones. It was found in Häggeby, ca. 100 km north-northwest of Stockholm. It indicates that
such depictions were not solely restricted to Gotland. The depictions do differ somewhat. On
Gotland the rowing vessels are always, as far as can be discerned, steered with two steering
oars and have some kind of construction in the midships. The vessel from Häggeby possibly
has a side rudder depicted. Montelius (1906) believed it to be somewhat younger than Nydam
B (p. 197).
Figure 24. The Häggeby stone. After Montelius, 1906, p. 197.
Chapter 4. The evidence: 4.2 Iconography 33
4.2.3 The Karlby pebble
A round pebble was found in 1987 on a beach near Karbly in Denmark. It measures only 12
by 7-8 mm. On it a vessel with a sail is carved on one side and on the other an animal with
horns (stag, reindeer or elk?). A photo and a drawing was published by Rieck and Crumlin-
Pedersen (1988) (see figure 25 and 26).
Figure 25. Photo by Djurslands Museum. After Rieck & Crumlin-Pedersen, 1988, p. 129.
Figure 26. Drawning by Werner Karrasch. After Rieck & Crumlin-Pedersen, 1988, p. 133.
Another drawing was produced by Hans Drake with the ship and animal drawn on the same
side (done for ease and not a depiction of the actual situation). Westerdahl (2015) remarks
that this drawing is based on two prototype drawings by Karrasch (p. 43). There are however
some differences observable when comparing to the drawing published by Rieck and
Crumlin-Pedersen and the one below. The most obvious is the absence of the vane on top of
the mast.
Figure 27. Drawing by Hans Drake. After Westerdahl, 2015, p. 43.
Chapter 4. The evidence: 4.2 Iconography 34
The pebble was found simply lying on the beach and could therefore not be dated by its
context. We are therefore restricted to the dating of the pebble by the figures alone. This
obviously results in different dates based on a researcher´s own interpretation of the figures.
Several have mentioned that it resembles Nydam B and Suttton Hoo (Haywood, 2006, p. 94;
Rieck & Crumlin-Pedersen, 1988, p. 130). Due to its “pre-viking” shape it has been given a
possible date to the 7th century (Crumlin-Pedersen, 1990, p. 111). Westerdahl (2015) has
given a date to ca. 800 AD (p. 43). Jones (1996) does not see any of the aforementioned
resemblance and believes that it rather resembles the vessels depicted on the Dorestad coins
(p. 84).
Figure 28. Examples of Dorestad coins. After Lebecq, 1990, p. 88.
It is clear that the dating is subjective and relies on the comparison to other shipwrecks and
other representations of ships. It is unknown for how long vessels with the Nydam shape have
continued to exist. Basing a date on the Nydam or Sutton Hoo II (630 AD) vessels is therefore
very uncertain. The combination of a horned animal and a ship seems also to be a long-lasting
tradition as the combination is present on the type B Gotlandic picture stones (Endre Skog,
figure 16) and on coins found in Birka, confusingly called the Hedeby coins, from the 8th
century.
Figure 29. A Hebey coin with on one side a ship and on the other a horned animal. After Arbman,
1940, Taf. 142, 9.
Chapter 4. The evidence: 4.2 Iconography 35
4.2.4 The Weser Bone
In 1927 a worked bone of a horse was found together with other bones at the Weser river near
Oldenburg in Germany. Many bones were found with runic inscriptions, but besides runes
this one also had a two-masted vessel carved into it (Göttlicher, 1989).
Figure 30. The Weser ship carving. After Göttlicher, 1989.
It was at first believed that the bones were fake, but it was later proven in a publication by
Pieper from 1989 that some were genuine including the one with the carving. It has been
interpreted as a Roman merchant ship. The runes read LOKOM:HER. Göttlicher (1989)
mentions that this probably means ´look here`, whilst Pieper (1989) translates it as “I'm lying
(look / see / watch) here” (p. 154). The ´I´ referring to the ship. The bone is dated to 310-450
AD on the basis of carbon-14 and amino-acids dating (Pieper, 1989, p. 241).
4.2.5 The Hvirring razor
On a razorblade dating to the Bronze Age, found in Honum, Denmark, a ship is represented
with a mushroom shaped extension (see figure) (Halldin, 1952, p. 75). Several have
interpreted this as possibly representing a mast and sail (Bugge, 1923, p. 29; Engelhardt,
1865, p. 15; Müller, 1897, pp. 445–446), others are sceptical (Farrell, 1979; Halldin, 1952, p.
74; Rieck & Crumlin-Pedersen, 1988, p. 156). Rieck and Crumlin-Pedersen state that it
“undoubtedly represents a solar disk as known from the Trundholm sun chariot” (1988, p.
156). Other interpretations have been that it represents an axe or a symbol of heaven (Boel
Bengtsson, 2017, p. 61). Some have interpreted it as representing a tree (Marstrander, 1976, p.
21).
Figure 31. The Hvirring razor. As published by Müller (left) and Worsaae (right). After Müller, 1897,
p. 444 and Worsaae, 1859, p. 36.
Chapter 4. The evidence: 4.2 Iconography 36
4.2.6 The vendsyssel vessel
In a bog in the Jetsmark parish, Northern-Denmark, a piece of wood with a 39 cm long vessel
carved into it was found (Marstrander, 1976, p. 461). It is 14C dated to ca. 1700 BC (Nancke-
Krogh, 1987, pp. 3, 6). Nancke-Krogh (1987) interprets the depiction as a skin boat with a
stem fashioned as the head of an animal. A vertical line is carved upwards from the vessel.
The slanting line to the right is a modern damage (ibid., p. 6). Nancke-Krogh states that it
cannot be a mast, but rather a stylised person. Any explanation why it cannot be a mast is not
given. Later on in the text it is stated that the earliest evidence for sailing vessels is that of the
Gotlandic picture stones and Kvalsund II (ibid., p. 13). It is therefore implied that the
possibility of it representing a mast is excluded because the first evidence is of a much later
date. With such reasoning an earlier depiction of a mast and/or sail will never be
acknowledged. Such circular reasoning should be avoided. Bengtsson (2017) is of the opinion
that it represents a mast just forward of midship (pp. 52-53).
Figure 32. The carving of the vendsyssel vessel. After Bengtsson, 2017, p. 53.
4.2.7 Rock art vessels with masts and/or sails
Within the mainstream discourse most researchers have stated that there are no certain
depictions of masts and sails on Bronze Age rock art or that these only depict vessels that
were paddled (Brøgger & Shetelig, 1950, p. 50; Kaul, 2003a, p. 187; Montelius, 1906, p. 125;
Rieck & Crumlin-Pedersen, 1988, p. 52; Vogel, 1915, p. 36) However, there are quite a few
rock carvings, dating to the Bronze Age, that have been interpreted by other researchers as
bearing a mast and/or sail (Bengtsson, 2017, pp. 59–60; Bugge, 1923, p. 29; Dahlgren, 1932,
pp. 46–48; Farrell, 1979; Halldin, 1952, pp. 70–78; Rausing, 1984, pp. 70–71; Stölting, 1996,
1999, p. 273). The carvings are too numerous to all be presented in this thesis, therefore only
a few are presented here. These are all presented in figure 33 and 36. For a complete overview
of carvings that can be interpreted as bearing a mast and/or sail and previous researchers who
have interpreted them as such, see Bengtsson (2017, pp. 59–87).
Coll (1902) seems to be the first to identify rock art vessels as bearing a sail (the ship from
Haugen, fig. 36, H) (pp. 45-46). Afterwards Dahlgren (1932) interpreted two vessels bearing a
mast from the Himmelstalund site (fig. 36, D & E). Although the interpretation of figure 36 D
is uncertain (Boel Bengtsson, 2017, p. 64). In 1952 Halldin made an overview of the rock art
vessels that until then were interpreted as bearing a mast and/or sail (Bengtsson, 2017, p. 60;
Halldin, 1952, pp. 70–78). In his work he includes Haugen, Solberg övre, Lilla Gerhum,
Chapter 4. The evidence: 4.2 Iconography 37
Harvaland and Järrestad-13, one boat from Himmelstalund, Vese Bro, Möckleryd (fig. 36,
E,F,G,H,I, J, N & O), another boat from Järrestad and one from Begby, Norway (Halldin,
1952, p. 77). Afterwards, in the early 1970´s, Göran Burenhult presented these and other
figures, identifying in total 38 figures of vessels with mast and/or sail (Boel Bengtsson, 2017,
p. 61).
During the end of the 1970´s and at the start of the 80´s, there were a series of notes regarding
the depictions of masts and sails from the Bronze Age in the Mariner´s Mirror. Farrell (1979)
refers to a figure published by Marstrander (1963, Pl. 59), a rock carving (Hafslund III) that
Farrell interprets as a vessel with a square sail (see figure 33). Farrell (1979) also refers to two
other depictions (Jaegersborg Dyrehave and Fruestraede, Denmark in Glob, 1969, pp. 25–26)
but finds these less convincing. Von der Porten (1978) also published two illustrations from
Southern Norway. One of them, from Begby, has also previously been proposed by Coll
(1903, p. 120). Von der Porten doesn´t find any of the by him given examples as representing
a mast and sail. Schovsbo (1980) is critical and of the opinion that all these examples must be
rejected (p. 16).
Figure 33. The carving from Hafslund III. After Marstrander, 1963, Pl. 59.
The carvings have received little attention in the mainstream discourse (Bengtsson, 2017, p.
67). When mentioned it was mainly to point out that these are not convincing and rather
depict other figures, such as a sun symbol or a standing figure (Rieck & Crumlin-Pedersen,
1988, p. 52). More recently the brothers Bengtsson have argued in several publications that
these depictions do represent vessels employing a sail (Bengtsson, 2017; Bengtsson &
Bengtsson, 2007, 2011). The earliest depiction of a vessel with mast and sail, according to
Boel Bengtsson, is the Tanum-833 from Bohuslän, Sweden (fig. 36, A). Bengtsson (2017)
seems to have dated it to 1500-1300 BC (p. 87). Other early examples are the Askum-44
vessel (fig. 34) and the large vessel at Himmelstalund (fig. 36, B). At first the Askum-44
vessel was dated to the early iron age by Humbla and Von Post (1937, pp. 70–72). Bengtsson
(2017), based on a scanned version of the image, interprets it as representing a Bronze Age
boat (fig. 35), dating it to Montelius period III (1300-1100 BC) following Kaul´s boat
chronology (p. 70).
Chapter 4. The evidence: 4.2 Iconography 38
Figure 34. Askum-44. To the left a photo and to the right the interpretation made. After Humbla & von
Post, 1937, p. 71.
Figure 35. Drawing of Askum-44 as interpreted by Boel Bengtsson. After Bengtsson, 2017, p. 67.
The depictions of possible masts and sails so far found mainly belong to Montelius period III
and IV (1300-900 BC) (Bengtsson, 2017, p. 87). Two possible figures from a later date are
the Harvaland vessels (fig. 36, F) and the vessel from Stjørdal (fig. 36, L). The Harvaland
carving was made on a slab of stone. Bengtsson dates the Harvaland ships to the transition
from the Pre-Roman (500 BC – 1 AD) to the Roman IA (1 – 400 AD) after Sognnes´ boat
classification (Bengtsson, 2017, p. 71; Sognnes, 2001). The Stuberg I boat, found in Stjørdal,
Norway, was dated by Sognnes to around 400 AD (Bengtsson, 2017, p. 73; Sognnes, 2001).
What is of importance to establish is whether the vessels with sails depict local vessel or not.
Several scholars are of the opinion that these depictions are copies of depictions on objects
that came to Scandinavia or through direct contact with the Mediterranean (Bengtsson, 2017,
p. 63). The possible depictions of vessel with a sail and/or mast are mainly found in the South
of Norway and Sweden, are found on different boat types and span over a long period
(Bengtsson, 2017, pp. 79–80). It seems to indicate that these possible sails and masts are
depicted on local vessels. This depends of course on which of the depictions are excepted. It
may be argued that the clear majority of the vessels depicted in the rock art do not have a
mast and sail depicted which therefore indicates only a sporadic depiction and indicates that it
was influenced by foreign objects. In the rock art there is however a general trend of not
depicting the means of propulsion. The usage of paddles is only depicted a few times
(Bengtsson, 2017, p. 81; Halldin, 1952, p. 76).
Chapter 4. The evidence: 4.2 Iconography 39
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Figure 36. An overview of several depictions with a possible mast and/or sail. A. Tanum-833, Sweden
after Bengtsson, 2017, p. 70; B. Himmelstalund, Sweden after Bengtsson, 2017, p. 69; C. Askum-44,
Sweden after Bengtsson, 2017, p. 67; D. Himmelstalund, Sweden after Bengtsson, 2017, p. 60; E.
Himmelstalund, Sweden after Bengtsson, 2017, p. 60; F. Harvaland, Norway after Halldin, 1952, p. 77; G.
Lilla Gerhum after Halldin, 1952, p. 77; H. Haugen, Norway after Bengtsson, 2017, p. 61; I. Solberg övre,
Norway after Bengtsson, 2017, p. 61; J. Järrestad-13, Sweden after Bengtsson, 2017; K. Herrebro-59,
Sweden after Bengtsson, 2017, p. 70; L. Stuberg I, Norway after Sognnes, 2001, p. 51; M. Torham-11,
Sweden after Bengtsson, 2017, p. 73; N. Vese Bro, Sweden after Bengtsson, 2017, p. 61; O. Möckleryd,
Sweden after Bengtsson, 2017, p. 61; P. Västra Tollstad-21, Sweden after Bengtsson, 2017, p. 77.
Chapter 4. The evidence: 4.3 Textual evidence 40
4.3 Textual evidence
Several historical texts have been used as evidence by scholars (see references) to prove or
disprove the presence of the sail. Only those texts that directly mention the use or absence of
the sail will be discussed here. These are;
- Tacitus´ Germania (Bengtsson, 2017, pp. 1, 48; Bill, 1997, p. 184; Dahlgren, 1932, p.
75; Haywood, 1991, pp. 11, 15, 19–21; Humbla & von Post, 1937, p. 70; Johnstone,
1980, p. 117; Rieck & Crumlin-Pedersen, 1988, p. 78; Vogel, 1915, p. 61).
- Gildas´ De Excidio Britanniae (Jones, 1996, p. 93; Vogel, 1915, pp. 67–68).
- Sidonius Appolinaris´ Letters (Bill, 1997, p. 184, 2010, p. 28; Crumlin-Pedersen,
1990, p. 113; Green, 1988, p. 54; Haywood, 1991, p. 71; Jones, 1996, pp. 75, 91;
Vogel, 1915, p. 67).
- Procopius´ History of the Wars, book VIII (Bengtsson, 2017, p. 48; Haywood, 1991,
p. 70; Jones, 1996, pp. 93–94; Lebecq, 1990, p. 85; Rieck & Crumlin-Pedersen, 1988,
p. 127).
4.3.1 Tacitus´ Germania
4.3.1.1 The Text
Translation by Mattingly:
Passing the Lugii, we find the Gotones under the rule of kings: a rule slightly stricter
than among the other Germanic peoples, but not yet beyond the bounds of freedom.
Then, on the Ocean, are the Rugii and Lemovii. All these peoples are distinguished by
round shields, short swords and submission to regal authority.
Next are the states of the Suiones, amidst the Ocean itself, which are strong not only in
arms and men but also in fleets. The shape of their ships differs from the norm in
having a prow at both ends, which is always ready to be put in to shore. They do not
rig sails or fasten their oars in banks at the sides. Their oarage is loose, as on some
rivers, and can be shifted, as need requires, from side to side. Wealth, too, is held in
high honour, and that is why they obey one ruler, with no restrictions and a claim to
compliance that cannot be questioned. Arms are not, as among the other Germani,
allowed to all and sundry, but are kept under custody, and the custodian is a slave.
There are two reasons for this: the Ocean makes sudden invasions impossible, and
armed men with nothing to do readily run riot. And of course it is not in the king´s
interest to assign control of arms to a noble or free man, or even a freedman. (Tacitus
Germania 44, transl. Mattingly, 2010)
Translation by Peterson of the sentences of interest:
Beyond these tribes the states of the Suiones, not on, but in, the ocean, possess not
merely arms and men but powerful fleets: the style of their ships differs in this respect,
that there is a prow at each end, with a beak ready to be driven forwards; they neither
work it with sails, nor add oars in banks to the side: the gearing of the oars is detached
as on certain rivers, and reversible as occasion demands, for movement in either
direction. (Tacitus Germania 44, transl. Peterson, 1914)
Chapter 4. The evidence: 4.3 Textual evidence 41
4.3.1.2 The reliability
Tacitus was born in 56 AD, possibly in southern Gaul. By the year 75 he commenced on his
career in Rome and in the year 97 he achieved the consulship. During this year he wrote
Agricola and the year after Germania. Agricola had a clear aim, to give tribute to his father-
in-law by compiling some of his greatest achievements, one of which was his role in the
conquest of Britain. Germania does not have such a clear aim and it is unclear why Tacitus
wrote this piece that best can be described as an ethnography. It describes the people of
Germania, the world beyond the limes.
Some have proposed that it was written to inform the emperor Trajan, then Governor of
Upper Germania, of the habits and military ways of the people in the North to facilitate the
conquest of Germania (Tacitus, Mattingly, & Rives, 2010, p. xxxix). Tacitus´ support for the
Roman conquest was very clear in Agricola, with his critique on the abandonment of Britain
by emperor Domitian (Tacitus et al., 2010, p. xxx). Another possibility is that Tacitus wanted
to comment on several societal issues by comparing the Roman people with those of
Germania (ibid., p xl-xli). If indeed it was the purpose to inform Trajan there doesn´t seem to
be any reason to alter the information that Tacitus received. For the other possibility there
neither seems to be any reason to distort the ´truth´ regarding the vessels used by these people
looked like and their method of propulsion.
Tacitus may have served on the Germanic frontier and would have had the opportunity to
gather information in these areas. Information regarding the people further north would likely
not have been first-hand observations. The information may have come directly from these
people, through long distance traders or auxiliaries, or it may have gone through a multitude
of people (ibid., p. xliii). It is impossible to assess the reliability since the original sources are
unknown to us.
4.3.2 Gildas´ De Excidio Britanniae
4.3.2.1 The Text
Translation Winterbottom:
Then all the members of the council, together with the proud tyrant, were struck blind;
the guard – or rather the method of destruction – they devised for our land was that the
ferocious Saxons (name not to be spoken!), hated by man and God, should be let into
the island like wolves into the fold, to beat back the peoples of the north. Nothing
more destructive, nothing more bitter has ever befallen the land. How utter the
blindness of their minds! How desperate and crass the stupidity! Of their own free will
they invited under the same roof a people whom they feared worse than death even in
their absence ´the silly prince of Zoan´, as has been said, ´giving foolish advice to
Pharaoh´.
Then a pack of cubs burst forth from the lair of the barbarian lioness, coming in three
keels, as they call warships in their language. The winds were favourable; favourable
too the omens and auguries, which prophesied, according to a sure portent among
them, that they would live for three hundred years in the land towards which their
prows were directed, and that for half the time, a hundred and fifty years, they would
repeatedly lay it waste. (Gildas De Excidio Britanniae 23, transl. Winterbottom, 1978)
Chapter 4. The evidence: 4.3 Textual evidence 42
Translation Giles:
A multitude of whelps came forth from the lair of this barbaric lioness, in three cyuls,
as they call them, that is, in three ships of war, with their sails wafted by the wind and
with omens and prophecies favourable, for it was foretold by a certain soothsayer
among them, that they should occupy the country to which they were sailing three
hundred years, and half of that time, a hundred and fifty years, should plunder and
despoil the same. (Gildas De Excidio Britanniae 23, transl. Giles, 1841)
4.3.2.2 The reliability
The words that are of importance from the Latin text are ´secundis velis´. The exact
translation, as can be seen above, is ambiguous. Winterbottom translates it as ´the winds were
favourable´ while Giles translates it as ´their sails wafted by the wind´. Vogel (1915)
translated it as possibly meaning favourable wind or swollen sails (pp. 67-68). More recently
Jones (1996) mentions that this may be translated as favourable wind, favourable sail or
smooth sailing, but that they do not refer to actual sails because the words are written in
connection with omens and prophecies (p. 93).
Gildas was born in the Kingdom of the Clyde, schooled in south Wales and later on migrated
to southern Brittany (Gildas & Winterbottom, 1978, p. 3; Jones, 1996, p. 122). He would
therefore have been well acquainted with the Saxons that lived in Britain. The text was
written sometime during the first half of the sixth century. Gildas clearly shows contempt
towards the Saxons. There seems no reason to state that they had sails if they did not use
those. However, it is unsure whether Gildas referred to a sail. What does become clear is that
the text is written in a highly literate fashion. It may therefore be that it is simply a literary
expression rather than an actual description.
4.3.3 Sidonius Appolinaris´ Letters
4.3.3.1 The Text
Translation Anderson, 1945:
I spent some hours with him in conversation about you, and he constantly affirmed
that you had recently sounded the trump of war in the fleet and, in discharging the
duties now of a sailor, now of a soldier, were roving the winding shores of Ocean to
meet the curving sloops of the Saxons, who give the impression that every oarsman
you see in their crew is a pirate-captain-so universal is it for all of them
simultaneously to issue orders and obey orders, to teach brigandage and to learn
brigandage. Even now there has cropped up a very strong reason for warning you to be
specially on your guard against danger from them. That enemy surpasses all other
enemies in brutality. He attacks unforeseen, and when foreseen he slips away; he
despises those who bar his way, and he destroys those whom he catches unawares; if
he pursues, he intercepts; if he flees, he escapes. Moreover, shipwreck, far from
terrifying them, is their training. With the perils of the sea they are not merely
acquainted-they are familiarly acquainted; for since a storm whenever it occurs lulls
into security the object of their attack and prevents the coming attack from being
observed by victims, they gladly endure dangers amid billows and jagged rocks in the
hope of achieving a surprise. Moreover, when ready to unfurl their sails for the voyage
home from the continent and to lift their gripping anchor from enemy waters, they are
Chapter 4. The evidence: 4.3 Textual evidence 43
accustomed on the eve of departure to kill one in ten of their prisoners by drowning or
crucifixion, performing a rite which is all the more tragic for being due to superstition,
and distributing to the collected band of doomed men the iniquity of death by the
equity of the lot. (Sidonius, Letters, book VIII letter VI To his friend Namatius, transl.
Anderson, 1965)
Translation Dalton, 1915:
I whiled away some time talking with him about you ; and he was very positive that
you had weighed anchor, and in fulfilment of those half military, half naval duties of
yours were coasting the western shores on the look-out for curved ships; the ships of
the Saxons, in whose every oarsman you think to detect an arch-pirate. Captains and
crews alike, to a man they teach or learn the art of brigandage ; therefore let me
urgently caution you to be ever on the alert. For the Saxon is the most ferocious of all
foes. He comes on you without warning ; when you expect his attack he makes away.
Resistance only moves him to contempt ; a rash opponent is soon down. If he pursues
he overtakes ; if he flies himself, he is never caught. Shipwrecks to him are no terror,
but only so much training. His is no mere acquaintance with the perils of the sea ; he
knows them as he knows himself. A storm puts his enemies off their guard, preventing
his preparations from being seen ; the chance of taking the foe by surprise makes him
gladly face every hazard of rough waters and broken rocks.
Moreover, when the Saxons are setting sail from the continent, and are about to drag
their firm-holding anchors from an enemy's shore, it is their usage, thus homeward
bound, to abandon every tenth captive to the slow agony of a watery end, casting lots
with perfect equity among the doomed crowd in execution of this iniquitous sentence
of death. (Sidonius, Letters, book VIII letter VI To his friend Namatius, transl. Dalton,
1915)
4.3.3.2 The reliability
Sidonius was born ca. 430 AD in East France. It seems that he has lived his entire life in
southern Gaul. During his time in Bourdeaux in 476 AD he met Saxons in the court and gives
a short description of them (Flierman, 2017, pp. 44–45). The text above might be a first-hand
description, but even if it was, Flierman (2017) warns that this is no guarantee for truthfulness
(p. 45). He states that Sidonius is still a Roman aristocrat with an aversion to barbarians. As
becomes also clear from the last sentence of the text above. The letter itself was composed ca.
480. Sidonius published several books of his letters but is clear that these are not the actual
letters he wrote but are modified for publication.
Chapter 4. The evidence: 4.3 Textual evidence 44
4.3.5 Procopius´ History of the Wars
4.3.5.1 The text
Procopius writes a story about a conflict between the nation of the Varni and the Angili from
Brittia. Radigis, son of Hermegisclus, king of the Varni (a people living north of the Rhine),
was to marry a maiden from Brittia, whose brother was king of the Angili. However,
Hermgisclus, knowing that he would soon die as was told to him by a bird, advised his son to
marry his step-mother, sister of the king of the Franks, since an alliance with the Franks was
by Hermegisclus deemed more important than with the Angili. After his father´s death,
Radigis did as commanded. This did however not fell well with the maiden from Brittia, who
therefore went to war against the Varni. She collected four hundred ships and put on board an
army of one hundred thousand men. It is this fleet to which the following sentences refer.
Translation Dewing:
And there were no supernumeraries in this fleet, for all the men rowed with their own
hands. Nor do these islanders have sails, as it happens, but they always navigate by
rowing alone. (Procopius, History of the Wars, book VIII, chapter XX, section 31,
transl. Dewing, 1962)
4.3.5.2 The reliability
The text was written during the sixth century. These sentences refer to the Angles. Procopius
does not seem to have been in direct contact with any of these people. Procopius source is
possibly a Frankish embassy who travelled to the emperor Justinian in Byzantium (Haywood,
1991, p. 70). The information written down is thus at least a second-hand description. For
Procopius himself there doesn´t seem to be any reason to alter the ´truth´. It is however the
question if the information of the informers can be trusted. The king of the Franks sent some
Angles with the embassy to “thus seeking to establish his claim that this island was ruled by
him.” (Procopius & Dewing, 1962, p. 255). If the information indeed comes from this
embassy they may have portrayed the Angles differently, as being of lesser status than the
Franks. This may be a reason to assert their ignorance of horses and possibly also of sails.
45
Chapter 5. The Arguments
As mentioned (section 3.3) the following will only focus on the arguments that are directly
based on the evidence that is presented in Chapter 4. Arguments based on other theories are
excluded.
5.1 The vessel remains
5.1.1 The Arguments
The construct of the argument deduced from the ship remains mainly relies upon the absence
of evidence, i.e. the absence of evidence that a vessel has been sailed. This is mainly based on
three vessels; Hjortspring, Nydam B and Kvalsund II. The first find that undoubtedly was a
sailing vessel is Oseberg dated to ca. 820 AD. Schematically the argument is as follows:
T
P1: None of the vessels found prior to 820 have been sailed
P1P1: Hjortspring, Nydam B and Kvalsund II have not
been sailed
P2P1: Oseberg has evidence of
having been sailed
P1P1P1:
Hjortspring has
no evidence of
having been
sailed
P2P1P1:
Nydam B has
no evidence of
having been
sailed
P3P1P1:
Kvalsund II has
no evidence of
having been
sailed
P1P2P1:
Has mast-step, mast, etc.
Prior to assessing the relevance of the premises it is important to assess their validity. The
question that should be answered is therefore: Do Hjortspring/Nydam B/Kvalsund II indeed
have no evidence of having been sailed?
5.1.2 What is looked for?
It is rarely explained what one exactly looks for when trying to determine whether a vessel
has been sailed. It is simply stated by the excavators that there is no evidence supporting that
a vessel has been sailed (Engelhardt, 1865, p. 13; Shetelig & Johannessen, 1929, pp. 58, 65),
which is thereafter repeated by others (Åkerlund, 1963, p. 22; Bill, 2010, p. 28; Brøgger &
Shetelig, 1950, p. 52; Brøndsted, 1940, p. 236; Dahlgren, 1932, p. 75; Montelius, 1906, p.
196; Nicolaysen, 1882, p. 11; Shetelig & Johannessen, 1930, p. 15). Engelhardt (1865) does
state that no masts were found and neither any facilities to attach ropes to, but it is still
unclear what is exactly meant with ´facilities´ (p. 13). In the case of Hjortspring, the
possibility of it having been sailed has never been discussed before Bengtsson and Bengtsson
(2011), only its paddling capabilities (Rosenberg, 1937, p. 89; Trakadas & Crumlin‐Pedersen,
2003).
Chapter 5. The arguments: 5.1 The vessel remains 46
In the case of Kvalsund II it is somewhat peculiar. Shetelig mentions several times the
remains of a possible mast (see section 4.1.3.5), however Johannessen states that “there is
nothing present that can inform us whether it Kvalsund II used a sail” (own translation)
(Shetelig & Johannessen, 1929, p. 65). Johannessen does state that it is possible, based on the
constructional features, that it used a sail, but that it is simply not possible to determine (ibid.,
p. 65). The latter statement has been referred to by several authors, but the focus so far has
mainly been on the absence of a mast or any such arrangements (Bengtsson, 2017, p. 49;
Christensen, 1984, p. 130; Greenhill, 1976, p. 187; Johnstone, 1980, p. 117; Sawyer, 1962, p.
76; Solberg, 2003, p. 207; Westerdahl, 2015, p. 48). Because of the absence of evidence of
mast and sail it is sometimes explicitly stated that a vessel was purely a paddling or rowing
vessel (Bill, 2010, pp. 27–28; Brøgger & Shetelig, 1950, p. 52; Engelhardt, 1865, p. 13;
Jones, 1996, p. 96; Nicolaysen, 1882, p. 11). Even when not stated it is most of the time
implied and further conclusion or theories are based on them not being sailing vessels.
During the excavation of the Nydam vessels no sailing vessels from the Viking Age had yet
been found. The well-known Tune, Gokstad and Oseberg ships were excavated in 1867, 1880
and 1904 respectively. These gave therefore an indication of what one should expect a sailing
vessel to look like prior to the excavation and interpretation of Hjortspring and Kvalsund II.
Thus the constructional features and rigging of these 9th and 10th century ships have become
the criteria for Scandinavian type sailing vessels (Bengtsson, 2017, p. 47). Additionally, the
Gotlandic picture stones show the earliest, widely accepted representations of vessels with
sails in the north. These finds seem to have given an a priori view of what should be looked
for. That is, a vessel with a square sail and the associated ´signs´, i.e. a keelson and mast-
support midships, facilities for fastening the running and standing rigging and the associated
constructional features (further discussed below).
5.1.3 Identifying a sailing vessel.
How can one identify a vessel as having been sailed? The difficulties regarding this question
have already been touched upon previously (see 2.1). If one wants to see whether a vessel has
been sailed one already needs to know what to look for. It is therefore likely that one is biased
due to preconceptions of what a vessel with a sail should look like. It is exactly therefore
necessary to clearly state what ´signs´ are looked for so that these may be criticised. A simple
statement that there was no evidence does not suffice. Due to this lack of clarity it is difficult
to assess the probability of the above premises. However, because of the mention of a
possible mast for Kvalsund II there may be some serious doubt. Several questions should be
asked prior to drawing any conclusions.
What kind of mast could have been used? It seems that there is a focus on finding a keelson or
mast partner (Bill, 2010, p. 28). Often it is implied that those should have been in place about
midships. Would a keelson and mast partner have been necessary? Other methods may also
have been used to attach a mast to a vessel and still transfer the forces on the mast to the
vessel (Bengtsson, 2017, p. 105; Ellmers, 1972, p. 98). Furthermore, mast-steps could also
have been placed in other constructional features than a keelson, such as a frame, and could
also have been placed in other locations if for example a bipod-mast was used. Although from
a later period (13th-16th century), a vessel (NZ 43) from the Flevopolder in the Netherlands
had a mast-step in chocks on port and starboard possibly indicating that it had a bipod mast
(Moortel, 1991, pp. 80-86,117; Reinders, 1985, p. 21). One should be open to several
possibilities (Bengtsson, 2017, p. 43).
Chapter 5. The arguments: 5.1 The vessel remains 47
What kind of sail could have been used? A square sail seems to be presupposed by many.
Likely because only square sails are depicted on the Gotlandic picture stones and Hedeby
coins. If the sail was present in Scandinavia earlier than accepted and it was an internal
development rather than an introduction, it is possible that different shapes have been
experimented with which eventually led to the square sail. Or it may be that sails were used
and after a period abandoned and later on re-invented or introduced. If there was a time of
experimentation it is also unlikely that immediately large sails were used. It is rather to be
expected that at first small sails were experimented with, irrespective of it being an internal
development or an introduction. Different types and sizes of sails can therefore be expected
(Bengtsson, 2017, p. 43). Kastholm (2007) has argued on the basis of the Gotlandic picture
stones that primarily low and wide ratio sails were used. Such a sail provides a lower centre of
effort, i.e. a vessel won´t capsize as fast as it would with a taller sail (Kastholm, 2007, p. 12).
Bengtsson (2017) is of the opinion that such sails are also indicated in the rock art (p. 91).
What kind of rigging could have been used and what kind of evidence would such rigging
leave on the vessel? For the attachment of the standing rigging one can simply fasten these to,
for example, the frames or the stems (McGrail, 1987, p. 219). Only some wear may then be
visible. It is the question what type of standing rigging was used. It is not even necessary to
use standing rigging. In southern Norway and northern Bohuslän, until quite recently, it was
not uncommon to sail without standing rigging, i.e. solely relying on the mast (Humbla & von
Post, 1937, p. 74). It seems that only if special features are present for the attachment of the
standing rigging, one is open to accept the possibility that it was sailed. This was the case for
the Ladby ship which had iron rings in its upper strake midship, possibly for the attachment of
the shrouds, which were the only remaining indications that it was sailed (Haywood, 1991, p.
67; Jones, 1996, p. 96). To use the absence of such signs as an argument that a ship was not
sailed, as for example Jones (1996) does for Sutton Hoo II, is not valid (p. 96). Even if special
attachments points were present it may be difficult to identify them as such. As mentioned the
cleats at the stems of Hjortspring have been interpreted as attachment points for a trussing
rope, but these can also be interpreted as attachment points for stays. The type of running
rigging used depends on what kind of sail was used.
A discussion on how to identify a sailing vessel is thus lacking in the discourse. Neither does
there seem to be a discussion on the preservation or the manner of deposition of the vessels
and how this may influence the likelihood of finding any traces of it having been sailed. It
seems that not much of Kvalsund II remained. What are then the chances that a mast-step was
preserved? Are other conclusions that have been drawn, e.g. that Kvalsund II had no beams,
justified? The conclusions made by early researchers seem to have been readily accepted by
others without a real discussion. It may be that only signs of wear indicate the use of a sail.
Any such marks may have been long gone when excavated. As mentioned only some
fragments were found of the keel of Hjortspring, so it may have had some sort of mast-step.
The possibility has never been discussed. Oseberg had facilities for rigging, but these were of
an organic nature and could have been long gone if it wasn´t for the unique preservation
(Bengtsson, 2017, p. 56; Haywood, 1991, p. 67).
Modifications prior to the deposition of a vessel also influences what eventually remains.
When a vessel served as a grave the body was normally placed in the centre and it may
therefore have been deemed necessary to remove the mast and keelson as has been pointed
out for the Sutton Hoo II (Ole Crumlin-Pedersen, 2010, p. 96; Haywood, 1991, p. 67). The
earliest vessels that do have the remains of a mast have a grave chamber placed at their ends.
The Kvalund vessels were ritually destroyed prior to their deposition. There may be several
Chapter 5. The arguments: 5.1 The vessel remains 48
reasons that specific parts of a vessel were treated differently than others, these parts may not
even have been deposited, deposited elsewhere or re-used (Bengtsson, 2017, p. 56;
Westerdahl, 2008, pp. 20–23, 26). There may be other, yet unknown, reasons for the absence
of certain parts besides preservation. Concluding that a vessel was not sailed because no mast-
step was found is very uncertain when the preservation is bad, but also when vessels were
found in a context where one may expect that ritual actions were undertaken, e.g. graves and
offerings. Concluding that a vessel was not sailed simply because no evidence was found does
not hold. Furthermore, it should be clearly stated what kind of signs are looked for.
5.1.4 Constructional features
That vessels prior to Oseberg did not use sails has been further argued for on the basis of their
constructional features. From the literature it becomes apparent that several features have
been taken as prerequisites for sailing vessels. One of them is an appropriate keel, by which is
meant a deep protruding keel for coping with leeway and one which is strong enough to
counteract the hogging and sagging forces of the ocean (Bengtsson, 2017, p. 56; Christensen,
1968, p. 24; Westerdahl, 1995, p. 42). A hull shape that is appropriate for sailing, i.e. not a
rounded cross-section but rather a more square, wider shape or wine-glass shape is preferred.
This also includes that the hull should be submerged for its entire length, i.e. no rocker or
raking stems (Christensen, 1984, p. 130; Ole Crumlin-Pedersen, 2010, pp. 70, 96). These
features seem to have been the most important, but additionally a high freeboard and a deep
rudder which is securely attached to the vessel were supposed to be present. Such criteria, as
discussed in section 2.1, can thus function as a list of criteria that a vessel has to fulfil in order
to be used for sailing. This is a valid method. The question is however if the above-mentioned
criteria are indeed required to be fulfilled for a vessel to be sailed.
Hjortspring and Nydam B are believed to not have been appropriate for sailing due to their
keel plank. It is questionable whether Hjortspring indeed had a keel plank (see section
4.1.1.3). The ruling thought is that a keel plank cannot cope with the forces of the sea
(sagging and hogging) and that a protruding keel needs to be in place in order to sail against
the wind and to resist leeway. As regards the forces exerted on the keel, the focus has clearly
been on oceanic voyages (Jones, 1996, p. 76). This is however not relevant for the use of
these vessels in more coastal protected waters. The reconstruction of the Hjortspring (Tilia)
could for example be used in waves of 1-metre high and Nydam B likely in rougher waters. A
trussing rope was used in Tilia for these sea trials, it is therefore not known how it would cope
without one. On the other hand, the thickness and the shape of the keel is neither known, so it
may also have been stronger then currently is believed. Whether the vessels could cope well
with the forces of the ocean is a different matter and not of relevance regarding their ability of
using a sail.
As regards the use of a keel for coping with the lateral leeway, this can also be solved by
other means (Bengtsson, 2017, pp. 42–43). This has been proven by the sailing experiments
of Tilia by Bengtsson (2017). No keel was needed for staying on course, the use of steering
oars at both ends of the vessel sufficed (Bengtsson, 2017, p. 112). The necessity for a deep
protruding keel is also less if a lower, wider and overall smaller sail is used. With such a sail
the Tilia was fitted and there was no need for a keel. It is thus possible to sail in a greater
wind force than would be possible with a high sail (Bengtsson, 2017, p. 103). It is thereby
also possible to sail against the wind, if deep enough steering oars or rudders are used or other
similar constructions. Tilia was capable of sailing at an angle of 90 degrees to the wind
Chapter 5. The arguments: 5.1 The vessel remains 49
without effort (Bengtsson, 2017, p. 112). With a canoe, lacking the same features of a more
advanced sailing vessel as Hjortspring and of a similar cross-section it was possible to sail
60-63 degrees to the wind (Bengtsson, 2017, p. 98).
The shape of the hull is also of importance for the leeway resistance. With a rounded hull, the
water passes easily underneath the hull and thus provides hardly any resistance. A squarer
shaped hull, provided it has sufficient draft, is not so easily moved sideways and thus resists
leeway. If the hull of a vessel has the shape of a wine glass as is the case with Gokstad, then
the lower part of the hull also functions as a keel as regards the leeway resistance. It is also
preferred that the vessel has no rocker and no raking stem, because if more length of hull is
submerged more of the hull works as a resistance to the leeway (Ole Crumlin-Pedersen, 2010,
p. 70). On the other hand, this can be partly solved with sufficient ballast which additionally
provides more stability.
Before one can assess the sailing capability of a vessel it is necessary to establish its shape
and construction. As could be seen in the overview of the evidence regarding the
reconstructions of the vessels there are uncertainties. Nonetheless, once a reconstruction
drawing is produced, conclusions are drawn about its capabilities which are used within the
discourse. The keel of the Hjortspring was only found in fragmentary pieces. It is therefore
not possible to draw any conclusions about the shape of its cross-section. Yet, the
reconstruction drawing made by Johannessen shows a flat keel plank and the Tilia has also
been constructed as such. For Nydam B several aspects were initially incorrectly reconstructed
most importantly its cross-section which results in quite some difference regarding its
capabilities. Common for these vessels is that reconstruction drawings have been produced by
Johannessen who was also responsible for the reconstruction drawings of Oseberg and
Gokstad. Christensen (1982) has pointed out the heavy reliance of Brøgger and Shetelig on
Johannessen who was a naval architect and mainly experienced with carvel vessels (p. 22).
Brøgger and Shetelig seemed to believe that the frames of the Viking ships were placed prior
to the planking, i.e. frame-first as in a carvel construction (ibid., pp. 22–23). Caution should
be taken with relying too much on Johannessen´s reconstructions. As regards the
reconstruction drawing of Kvalsund II by Johannessen, there seems to be a lot of uncertainty.
Whether the drawing come about through an objective method is questionable.
5.1.5 The evolutionistic view
On the basis of Hjortspring, Nydam B, Kvalsund II, Oseberg and Gokstad an evolutionary
development has been advocated by Shetelig. This was already apparent in several earlier
publications but was for the first time clearly formulated in Shetelig and Brøgger´s book from
1950. In it, the aforementioned vessels are put in an evolutionary line with Kvalsund II being
the first stage towards the development of the sailing ships of which Oseberg is the first
example culminating in the more advanced Gokstad. Other more fragmented finds have been
seen as further supporting this theory. Before the use of dendrochronology and 14C dating this
evolutionary line or typology, has been used to date finds, as was the case for the Kvalsund
vessels. It is indeed so that changes are visible in the general construction based on the current
finds., e.g. from sewing to the usage of nails and roves and from frames lashed to cleats to
frames fastened directly to the planks with treenails. However, making further generalising
statements can be quite problematic since it may lead to a simplistic representation of reality.
I believe that this is the case with the evolution of the Scandinavian vessels as advocated by
Brøgger and Shetelig. Scandinavia is often seen as a region with strong internal connections.
Chapter 5. The arguments: 5.1 The vessel remains 50
Ship finds from all over Scandinavia have therefore been set in an overall evolutionary chain
not taking into account the possibility of regional diversity. Logically, in times of close
contact, the vessels of others will be closely observed. However, it may be questioned
whether this leads to a standardisation of ship constructions. Accepting a find from Denmark
to be representative for a whole period, for the whole of Scandinavia is quite questionable
(Ejstrud & Maarleveld, 2008, p. 68). This is however what is done within the evolutionistic
view and within the discourse regarding the adoption of the sail.
Additionally, the possibility of vessels having different functions, i.e. the existence of
different types within the plank-built tradition is not taken into account. Thus, Nydam B and
Kvalsund II resulting finally in the Viking ships are seen as part of the same development
although these ships are from completely different environments and possibly had different
functions. By advocating such an evolutionary teleologic view any possibility of choice made
by the boatbuilders or intended user is ignored. What sense would it make to have a deep
protruding keel on a vessel in the shallow waters of Denmark with its changing sandy sea
bottom? Indeed, the Oseberg and Gokstad ships with their deep protruding keels were found
in the Norway, whilst the Skudelev vessels still have quite shallow keels (McGrail, 1987, p.
115). The first vessel with a T-keel is Kvalsund II (or possibly Nydam C depending on exact
definition) also found in Norway. What is mainly problematic for the discourse is that the
main reason for this development has been said to be the development towards the sailing
ships, i.e. that the boatbuilders of the North sought to transform their rowing vessels into
sailing vessels prior to the adoption of the sail.
Since the well-known Oseberg, Tune and Gokstad ships used a sail, thus displaying the end of
the alleged evolutionary chain, their constructional features have become the prerequisites for
Scandinavian type ships that were able to sail. This seems also partly due to the constant
focus on the migration of the Angles and Saxons and the sudden expansion of the Viking Age
which required vessels capable of crossing the open sea under harsh conditions. There is an
implicit reasoning that if a sailing vessel cannot make such crossings it was not capable of
sailing at all. This clearly does not take into account the vessels´ intended function and
environment. Making the assumption that during the migration period no ships were available
for crossing the North Sea on the basis of the Nydam vessels can neither be made. The
Nydam vessels have normally been seen as warships or troop transporters and were found on
the east coast of Denmark. It may be that the Nydam vessels were purposely made for the
sandy beaches, the calmer waters in contrast to the rougher North Sea and for having good
rowing capabilities. It can therefore be that it was chosen to not equip the Nydam ships with a
mast. The Saxons were well known for their surprise attacks, even to such an extent that they
would use a storm to hide their boat in (see section 4.3.3). Equipping a boat with a mast may
not be a logical choice if one wants to keep it hidden, especially for more local attacks. If it
was deemed not too far to row, it may have been more logical to use a boat that was less
visible, easy to land on a beach with its keel plank and fast under oars due to its rounded and
slender hull instead of a larger sailing vessel. Such a sailing vessel can be quite problematic
for sudden attacks. With a deeper protruding keel one can go less near the beach and it may
therefore take more time to attack and to make an exit. Although it may be more stable it is
likely slower under oars. When it is under sail it will be faster, but it is first necessary to row
the vessel until the wind is favourable. Especially in landscapes were this may be difficult,
such as in fjords, it may be more advantageous with a fast rowing vessel.
A vessel is made for a function and the environment in which it is going to be used. I will not
propose that the above reasons are why the Nydam vessels are designed such as they are. It is
Chapter 5. The arguments: 5.1 The vessel remains 51
rather to point out that their intended function(s) and environment may have been considered
for their design. Rather than conscious choices, the vessels are however seen as part of an
evolutionary chain and thereby representative for that period. It may be that they only used
one type of ship, an all-purpose vessel, but the possibility of diversity by choice has to be
taken into account before simply accepting them as representative for all the vessels from that
period.
5.1.6 Representativeness
Due to this evolutionistic view the representativeness is hardly discussed. When it is
discussed, it becomes clear that the vessels (Hjortspring, Nydam B and Kvalsund II) may not
be so representative and that other types of vessel may be overlooked. However, there is an
assumed standardisation of construction in Scandinavia, i.e. that the constructional features of
one vessel is transferable to others found in Scandinavia if they have a few corresponding
features and are from roughly the same period. This becomes clear from the following
example.
Jones (1996) remarks that the finds from the offerings and burials give a distorted view, since
it is unlikely that one would have used a cargo vessel for a high-status burial (p. 82). This was
also pointed out by Crumlin-Pedersen (1990) as he states that boats from “different context
tend to produce different types of ships and boats” (p. 98). It is therefore argued that only
fragments from other contexts, e.g. as wrecks or as re-used timber, will be representative for
the more common vessel. So far the reasoning seems to be logical with valid remarks. Jones
mentions the fragments of three vessels; the Gredstebro, Hjemsted and Kongsgårde. Jones
concludes that these fragments are “consistent with the overall pattern of shipbuilding
represented by the ship burials. Nor is it absolutely certain that ship burials involve only
special and therefore perhaps unrepresentative vessels” (p. 82). Problematic is that these
fragments give only a few indications of their construction. They do not provide any further
support for the absence of the sail. Of the Gredstedbro (7th century) vessel only 3 timbers
were taken to the museum although many more were found. These three timbers are a part of
the keel, a part of a stem that features a scarf and a parts of a frame. The entire cross-section
of the keel has not been preserved, but it could be deduced that it would have been quite
shallow (Ejstrud & Maarleveld, 2008, p. 63). Crumlin-Pedersen (1968) further stated that: ”It
was probably a rowing vessel, in any event its keel is not the characteristic keel of a sailing
ship” (p. 267). Of the Hjemsted vessel only one timber was found in a well. Rieck and
Crumlin-Pedersen (1988) state that it is rather a keel than a plank although it is unclear why
they think so since no further arguments are used (p. 135). As regards the Kongsgårde vessel,
only two large frames were found, thus any statement of its capability to use a sail or not
cannot be made. nevertheless, Crumlin-Pedersen finds the frames similar to that of Sutton
Hoo II and states about these two vessels that “the hull shape shows no signs of adaptation to
the requirements of a sailing ship hull such as a deep or long keel or sharp ends to prevent a
strong leeway” (Crumlin-Pedersen, 1990, p. 111). This transference of features from Sutton
Hoo II based on two frames cannot be made and is pure speculation. Arguing that it had a flat
keel since the frames are of a similar size and shape as those of the Sutton Hoo II is not more
than an assumption. Although fragments from other contexts are considered, their unknown
features are assumed based on already known vessel. The fragments are thereby seen as
evidence to support the standardisation of construction whilst their reconstruction is based on
this assumption, i.e. circular reasoning.
Chapter 5. The arguments: 5.1 The vessel remains 52
A recurrent remark throughout the discourse is the alleged conservatism of the Scandinavian
boatbuilders (Christensen, 1982, p. 22; Greenhill, 1976, p. 178; Jones, 1996, pp. 85, 95). The
idea that the Scandinavian boatbuilders were conservative, more than other boatbuilders, is
not based on evidence and seems misplaced. The only ´conservatism´ is their adherence to
building their vessels as lightly as possible, which is the basis for the Scandinavian
boatbuilding tradition. This seems to stem from their continuing reliance on various methods
of propulsion which are not solely reliant on wind. It seems therefore logical that they adhered
to their clinker-building tradition, it fulfilled their requirements. Even so, in my opinion, there
seems to be much change within this tradition. There are for example plenty of differences
between Hjortspring, Nydam B and Oseberg. I would not call this conservative, at least not
more conservative than other boatbuilding traditions. As discussed, there are clear problems
with the evolutionistic as regards the representativeness. Nydam B has for example been taken
as representative for the level of construction in Scandinavia at its dated time, but is this really
the case? That depends if there were other vessels during that period which were more
´advanced´. Nydam C has been said to have been more advanced in its construction. It might
as well be that the Nydam C was more representative for the level of construction for that
period. The Nydam C has however been dated to a later period because of its more advanced
construction. Thus the typology of the development of the Scandinavian vessels prohibits the
falsification of the idea that the boatbuilders in Scandinavia were conservative. Nonetheless,
on the basis of this ´conservatism´ it has been argued that the likelihood of an undiscovered
very different vessel is unlikely (Jones, 1996, p. 85). It is seemingly also often the best-
preserved vessel that is taken as representative for a period. Thus Nydam B is often seen as the
vessel that was used for the migration to England (Bengtsson, 2017, p. 50), whilst Nydam C is
hardly taken into account although it shows rather different constructional features.
Admittedly its reconstruction is uncertain, but according to Åkerlund Nydam C has a stronger,
more developed keel than Nydam B and is thus more suited for open waters. However,
conclusions are drawn solely on the basis of Nydam B.
As regards Hjortspring, vessels with a similar shape on rock art from the end of the Bronze
Age and Pre-Roman Iron Age have been found throughout Scandinavia. It may therefore be
representative for a common ship type in Scandinavia as stated by Kaul (2003a, p. 187).
However a similar side view does not have to imply a similar construction. A find from
western Norrland in Sweden of a similar upper crossbeam as that of Hjortspring has also been
used as evidence to support a wider geographical use of such vessels (Crumlin-Pedersen,
2003b, p. 221). Yet even if this is the case it cannot be stated this was also the case for later
periods. Several regions may have afterwards developed their own distinctive ways of
construction.
Chapter 5. The arguments: 5.1 The vessel remains 53
5.1.7 Assessment of the arguments
As mentioned the structure of the argument based on the ship remains is as follows:
T
P1: None of the vessels found prior to 820 have been sailed
P1P1: Hjortspring, Nydam B and Kvalsund II have not
been sailed
P2P1: Oseberg has evidence of
having been sailed
P1P1P1:
Hjortspring has
no evidence of
having been
sailed
P2P1P1:
Nydam B has
no evidence of
having been
sailed
P3P1P1:
Kvalsund II has
no evidence of
having been
sailed
P1P2P1:
Has mast-step, mast, etc.
The probability of the premises that Hjortspring and Nydam B have no evidence of having
been sailed has always been judged to be High (see 3.2 for terminology). As discussed it may
not be so simple to conclude that a vessel has not been sailed. As regards Hjortspring much
was fragmented and no keel remained. The cleats at the stems have been interpreted as
attachment points for a trussing rope while Bengtsson has interpreted it as possible fastening
points for stays. Even if facilities for rigging may remain it may therefore be difficult to
recognise them. There is thus uncertainty whether there were indeed no signs. These have at
least not been noticed and interpreted as such by their excavators. Since Hjortspring has
possible attachment point for stays I will judge the premise P1P1P1 as having an intermediate
probability. As regards Nydam B, no evidence of the use of sails has been identified and I
would therefore judge the premise P2P1P1 of a high probability. As regards Kvalsund II a
possible mast may have been found as mentioned by Shetelig. Therefore I can only assign
P3P1P1 a weak probability. I will judge the relevance of these premises as intermediate since
actions prior to the depositions and the preservation may give a distorted view and since it is
unclear which evidence, or signs, were looked for.
These premises influence the strength of premise P1P1. Especially P3P1P1 is a weak premise,
yet Kvalsund II is the youngest of the three and only moves the first evidence of the use of the
sail to ca. 700 AD, i.e. still being a late adoption. P1P1 has been further supported by the
criteria for a sailing vessel which Hjortspring and Nydam B do not fulfil and Kvalsund II
partly (lack of high freeboard). These criteria have been discussed and it is clear that the
absence of these does not exclude the possibility that a vessel used a sail. These give therefore
no further support for the premise. The probability of P1P1 is therefore weak because the
strength of the argument for Kvalsund II is weak. If Kvalsund II is not taken into account the
validity of P1P1 would be Intermediate since the strength of the argument regarding
Hjortspring and Nydam B is judged to be good.
The relevance of P1P1 for P1 is that Hjortspring, Nydam B and Kvalsund II are representative
for the vessels (fragments) found until now. These vessels are from different periods and as
such they may be deemed representative for quite a long period. What is problematic is that
Chapter 5. The arguments: 5.1 The vessel remains 54
these vessels were found in the same context, that is, as a bog offering. Their function, based
on this context, seems to be war vessels (mainly as troop transporters). Other vessels found
may however have had different functions. Since the remaining finds are rather fragmentary it
is difficult to judge, but through the evolutionary theory it is implied that they are similar to
the vessels discussed. It often seems to be the case that the vessels reconstructed from
fragments are influenced by what is already known from other finds rather than being
independently inferred from the fragments themselves. It is clearly not valid to use these
fragments accordingly to and based on the evolutionary theory and then conclude that they
therefore support the theory. This is an clear example of circular reasoning. As regards the
Nydam vessels, the focus has been mainly on Nydam B, although Nydam C shows advantages
for use in rougher water such as a stronger keel and a scarf from keel to stem which is also
stronger. Due to the problems of the representativeness of these vessels and because the
advocated constructional criteria are not necessary for a sailing vessel, the premise P1 is
judged to have a weak probability.
The strength of the relevance of P1 for the Thesis (T) is dependent on the representativeness
of the vessels found until now that have been confirmed to not have been sailed. As discussed
previously diversity has not been considered. Thus instead of discussing the possibility that
Hjortspring and Nydam vessels were constructed with a certain function and environment in
mind they are seen as part of an evolutionary chain. Furthermore, `Scandinavia´ is implied to
be one continues common region throughout the entire pre-history. Such an assumption
should be argued for, not simply accepted. It is therefore not possible to state that Nydam B is
representative for all of the vessels in Scandinavia during that period. Even if this is the case it
cannot be excluded, on the basis of its construction, that this type of vessel could not have
been sailed. Because of the above-mentioned problems with the current argumentation for the
representativeness I will judge the relevance for the Thesis as weak.
Chapter 5. The arguments: 5.2 Iconography 55
5.2 Iconography
5.2.1 Bronze/ early Iron Age iconography
The representation of vessels with a possible mast and/or sail from the Bronze Age and Early
Iron Age (the Hvirring razor, the vendsyssel vessel and rock art) are here treated as one since
the same argumentation is applied, which is as follows:
¬T
P1: Bronze/ early Iron Age iconography includes depictions of vessels with a mast and/or
sail
The argument seems simple, but in reality is complex. It can hardly be proven that the
depictions indeed show a vessel with a sail. It all comes down to a subjective interpretation.
However, the possibility cannot be disregarded. In the mainstream discourse this is
nonetheless happening. The interpretation of the Hvirring razor has been in discussion, with
intervals, for quite some time. Within the mainstream discourse it has for a long time been
seen as the only possible representation of a sail in Scandinavia even though several have
pointed out possible sails in rock art depictions (see section 4.2.7). These have received little
attention. When the rock art was mentioned it was mainly to assert that there were no
representations of sails. Any figure above a vessel has mainly been interpreted as a
representation of the sun. This has grown to be a default interpretation. This leaves not much
room for the possible interpretation that it represents a sail. Similar problems arise as with the
identification of a sailing vessel discussed previously. What should one look for? The first
accepted representation of sails are square sails which were used throughout the Viking Age,
but can one simply state that one should also expect a square sail in the Bronze Age? One has
therefore to be open for different shapes of sails and different locations of the mast or masts.
This may be a single mast, a bipod mast or multiple masts. When it is stated that one does not
believe it to represent a mast and/or sail one has to explain why not. An alternative
interpretation and its likelihood needs to be underpinned.
The scepticism towards the possibility of the iconography representing vessels with sails also
seems to stem from the general idea that such vessels could not cope with the forces of a mast
and sail. As previously discussed this is not the case. It is also the question what size of sail is
used. Most seem to imagine the use of a large sail as is the case for the Oseberg and Gokstad
vessels. This needs of course not be the case. A sail could also simply be used as a helping
device as an addition to paddling or rowing if the wind was favourable (Bengtsson, 2017, p.
121). The depiction of a sail may therefore be not as obvious as has been thought (Brøgger &
Shetelig, 1950, p. 50). The probability of the premise is therefore dependent on one’s own
interpretation. The possibility that sails are depicted can however not be excluded. Personally
I find some of the representations convincing and assign it therefore with an intermediate
probability. As for the relevance of the premise for the anti-thesis it is the question, if the
premise is accepted, whether that it can be concluded that the sail was indeed used in
Scandinavia. This is dependent on the acceptance that the vessels depicted are vessel that
were used in Scandinavia and not simply depictions of foreign vessels as has been proposed
for the Hvirring razor and rock art (Bengtsson, 2017, p. 63; Dahlgren, 1932; Vogel, 1915, p.
36). As mentioned (section 4.2.7) the depictions seem to be widespread, mainly in Southern
Norway and Sweden, and are from different periods. This seems to indicate the usage of the
Chapter 5. The arguments: 5.2 Iconography 56
sail on local vessels, that is, if the depictions from the different periods are accepted which
again is subjective. It also dependents on the strength and type of connections in the Bronze
Age with the rest of Europe, but this is outside the scope of this thesis. Based on my personal
opinion I will judge the relevance as strong.
The exact date of the representation of a sail is not crucial. No matter which depiction is
accepted this will be of an earlier date than currently accepted and thus a falsification of the
current thesis. The strength of the argument thus relies on the probability of the premise. It is
up to the reader to decide if the Bronze Age iconography represents a mast and/or sail.
5.2.2 The Karlby pebble
As discussed, the dating of the pebble is ambiguous. Crumlin-Pedersen (1990, p. 111)
proposes a date to the 7th century whilst Westerdahl (2015) dates it to ca. 800 AD (p. 43).
Haywood, on the bases that it resembles Nydam B, sees no reason why it could not date to the
4th century or earlier (1991, p. 21). The date of the Karlby pebble can therefore not be used as
evidence to support one or the other theory. Regardless of the date Haywood states: “if a
Nydam type ship could carry a sail in the seventh century, it could carry a sail in the fourth
century”. Such an argument does not hold. The Karlby pebble may resemble the shape of
Nydam B as seen from its side, but it cannot be concluded that Nydam B employed a sail. It is
not known for how long such vessels were continued to be used. Since no objective date can
be given to the stone it can hardly be used as evidence as has been pointed out before (Bill,
2010, p. 25). The probability of any premise or conclusion based on the pebble regarding the
adoption of the sail is therefore weak.
5.2.3 The Weser Bone
The Weser Bone has hardly been mentioned in the discourse. The only mention of it, as far as
I could find, is by Westerdahl (1995) and Haywood (2006, p. 35). The bone shows that the
Germanic people must have known of the square sail since it was found at the Weser and had
runic inscriptions on it indicating that the carving was made by someone in the area. The bone
was previously thought to be a falsification. This must be why it is hardly used as evidence
within the discourse. It was proven to be genuine in 1989 by Pieper and is dated to 310-450
AD. However, also after it was proven to be authentic it has hardly received attention. It has
not been used directly as an argument for an earlier use of sail, but it possibly shows that a
roman type vessel was seen in the Weser in the 4th/5th century. At least it supports the theory
that the Germanic people were acquainted with the use of the sail. Some have argued that it
would therefore be logical that the North-Germanic people also would have taken the sail into
use (Haywood, 2006, p. 35). However, the knowledge of new technology does not imply the
adoption of it. There may be several reasons why it was not adopted or not adopted
immediately. Such reasons may be societal, technological or ideological (Westerdahl, 1995).
Chapter 5. The arguments: 5.2 Iconography 57
5.2.4 The Gotlandic Picture Stones
The earliest evidence that is normally referred to for the use of the sail in Scandinavia is that
of the Gotlandic Picture Stones. The structure of the argument is as follows:
T
P1: The earliest depiction of a sailing vessel in Scandinavia dates to the 6th/7th century.
P1P1: The Gotlandic type B stones have
depictions of vessels with sails.
P2P1: No earlier depictions of vessels with
sails are depicted in Scandinavia.
The probability of premise P2P1 and thereby P1 can be contested by depictions of vessels with
sails from an earlier period such as those discussed previously. As mentioned this is a matter
of interpretation and it is not possible to determine whether these are indeed depictions of
vessels with sails. But there is reason to doubt P2P1. The premise can therefore not be
assigned a high probability. Another argument that can be formulated to support P2P1 is that
the change from rowing to sailing vessel on the picture stones reflects reality (Bill, 1997, p.
185). It is questionable whether such a conclusion can be accepted. Does the absence of a sail
on type A stones exclude the possibility that the sail was used? Here it is important to take
into consideration the intended function of the picture stones and the meaning of the
depictions. If it is connected with beliefs of the dead, and therefore religious and
mythological, what kind of vessel is portrayed? Is it a common, every day, type vessel or does
it represent a certain type of vessel specifically connected with the dead? Crumlin-Pedersen
(2010) has for example argued that the ship graves are connected to the Vanir cult and that the
vessels in the graves represent Skíðblaðnir, the ship-of-luck, of Freyr (pp. 155-156). One
should therefore take caution when interpreting these depictions as simply reflecting the type
of vessels which were commonly in use. Rather than displaying the change from rowing to
sailing vessels it might have been a change in symbolism or in function of the stones
(Bengtsson, 2017, p. 55). The type A stones do therefore not exclude the use of the sail.
Assessing the probability of P1 depends on the probability of the premise that there are earlier
depictions of sails. This is rather subjective and may either be given an intermediate or a weak
probability. Based on my personal interpretation I will assign P1 a weak probability since I
find some examples of the rock art rather convincing.
The relevance of P1 for the thesis (T) is that the change in the depiction of the Gotlandic
picture stones reflects a change in the vessels in use. This has also been assumed for the later
picture stones. It is believed that the Gotlandic picture stones reflect the evolutionistic theory
from rowing vessels to gradually more advanced sailing vessels (Christensen, 1982, p. 26). As
already argued this is questionable. The vessels depicted may be a special type of vessel or it
may be simply a change in style as has been proposed by Varenius (1992). Crumlin-Pedersen
(1965) has argued that the first sailing vessels represented were in fact Frisian ships who
introduced the sail to Scandinavia. This theory has its flaws and is based on several
assumptions. One objection is that the picture stones were found in connection to graves and
are believed to be of a religious and mythological character. The choice of depicting a foreign
vessel on such a stone, does not seem logical to me. The current thought seems to be that the
stones display ´Scandinavian´ vessel types. The iconography on the Gotlandic type B picture
Chapter 5. The arguments: 5.2 Iconography 58
stones show therefore the usage of the sail on vessels presumably from Gotland. It is only
possible to state that the Gotlandic picture stones tell us something about the people on
Gotland. Stating that the changes on Gotland are representative for the rest of Scandinavia
needs to be argued for, something which has not been done. As is the case for the remains of
the vessel, changes occurring at one location in Scandinavia are immediately seen as
representative for the whole of Scandinavia. Such a generalisation cannot be so simply
accepted. Thus if the stones indeed are a reflection of reality it cannot be generalised to
include Scandinavia. The relevance of P1 for the thesis is therefore weak.
Chapter 5. The arguments: 5.3 Historical evidence 59
5.3 Historical evidence
5.3.1 Tacitus
Although there seems to be no reason for Tacitus to distort the information he received, it is
questionable if the information itself was correct (see 3.3.1). Apart from this uncertainty what
does the text actually tell us?
The sentence that is of importance is the following: “They do not rig sails or fasten their oars
in banks at the sides.” (translation Mattingly) or “they neither work it with sails, nor add oars
in banks to the side” (translation Peterson). This statement regards the state of the Suiones,
which is generally accepted to be in Sweden. The statement either refers to the Suiones
themselves thus including all the vessels they used as can be inferred from the text translated
by Mattingly: “Next are the states of the Suiones, amidst the Ocean itself, which are strong
not only in arms and men but also in fleets. … They emphasis added do not rig sails or
fasten their oars in banks at the sides.” Or it may specifically refer to the vessels used for war
if the translation by Peterson is used: “Beyond these tribes the states of the Suiones, not on,
but in, the ocean, possess not merely arms and men but powerful fleets … they neither work it
emphasis added with sails”.
The description of the vessels given is reminiscent of the Hjortspring;
The shape of their ships differs from the norm in having a prow at both ends, which is
always ready to be put in to shore. … Their oarage is loose, as on some rivers, and can
be shifted, as need requires, from side to side. (translation Mattingly, 2010, p. 55)
The loose oarage seems to indicate the use of paddles. Some have also interpreted the
statement as describing Nydam B (Montelius, 1906, p. 196). However if the direction would
be shifted then the distance between the thwarts would be very irregular (Åkerlund, 1963, p.
56). Furthermore the oars were partly fastened with ropes and the oarlocks could not be so
easily shifted in direction. The text was written in 98 AD and it is therefore difficult to judge
what the vessels referred to looked like since it is ca. 400 years after Hjortspring and ca. 200
years before the Nydam vessels.
The statement by Tacitus has often been used as an argument against the usage of the sail in
Scandinavia (see section 4.3). Such an argument is based on two assumptions: the vessels and
method of propulsion of the Suiones are representative for the rest of Scandinavia and that the
vessel type described is the only type in use (if the translation by Peterson is followed)
(Bengtsson, 2017, p. 48). Haywood (1991) is of the opinion that Tacitus specifically
mentioning the Suiones as not having sails indicates that they were different from the other
Germanic people in this aspect (p. 21). This may be supported by Tacitus stating: “The shape
of their ships differs from the norm” (Mattingly); “the style of their ships differs in this
respect” (Peterson); “forma navium eo differt” (Tacitus). It is then the question in comparison
to whose vessels it differs. To me it seems that this refers to other Germanic groups, i.e. that it
is not a comparison to the vessels of the Romans.
Throughout Germania a specific feature of each tribe is pointed out in order to make apparent
how they differ from other tribes, i.e. how they can be recognised (see the beginning of the
passage, p. 42). One cannot on the basis of Tacitus´ statement conclude that no groups in
Chapter 5. The arguments: 5.3 Historical evidence 60
Scandinavia used a sail (Bengtsson, 2017, p. 48). At best this statement can give us
information regarding the Suiones.
The premise, that the Suiones or their war vessels had no sail, is unsure since the information
that Tacitus received may have been unreliable. Regardless of its reliability the statement can
hardly be generalised to include all the other people in Scandinavia. The relevance to the
thesis is therefore weak.
5.3.2 Gildas
The passage by Gildas has been used as an argument for an earlier use of the sail. Whilst
writing about the Saxons the following line is written: “Then a pack of cubs burst forth from
the lair of the barbarian lioness, coming in three keels, as they call warships in their language.
The winds were favourable; favourable too the omens and auguries” (translation
Winterbottom), “A multitude of whelps came forth from the lair of this barbaric lioness, in
three cyuls, as they call them, that is, in three ships of war, with their sails wafted by the wind
and with omens and prophecies favourable” (Translation Giles). The words that are of
importance in the original text are: secundis velis. The difficulties regarding this translation
have been already pointed out (see 3.3.2). Due to the problems regarding the translation it is
not possible to say whether it can be used as an argument for the use of the sail. If it only
means “favourable wind” it does not have to imply the use of a sailing vessel, since
favourable wind is favourable for any kind of vessel. Additionally the text by Gildas is highly
literary and it may just as well be a literary phrase rather than an actual description (Jones,
1996, p. 93). Due to the uncertainty of translation it is unreliable and its use as a premise
therefore weak. Furthermore, it is the question in which region the Saxons lived that is
referred to. Gildas may mainly have been familiar with the Saxons already living in Britain,
which may have been somewhat different to the Saxons living in their place of origin.
Whether this place of origin can be said to include ´Scandinavia´ is also the question, and thus
its relevance for the thesis is unsure. The main thought being implied seems to be that if the
Saxons used sails then the people further north also used them since they had similar
constructed vessels. That the Saxons used sail does however not mean that the people further
north also used the sail, but it may show that the constructional features that have been
discussed as prerequisites for the adoption of the sail were indeed not seen as prerequisites for
the Saxons and therefore likely neither for other people. The use of the sentence as a premise
is judged to be weak and due to it only discussing the ´Saxons´ the relevance for the Thesis
(T) is also judged to be weak.
5.3.3 Sidonius
The text by Sidonius also regards the Saxons. There are several sentences of relevance for the
discussion, which are:
- “Moreover, when ready to unfurl their sails for the voyage home from the continent
and to lift their gripping anchor from enemy waters”
- “give the impression that every oarsman you see in their crew is a pirate-captain-so
universal is it for all of them simultaneously to issue orders and obey orders”
Chapter 5. The arguments: 5.3 Historical evidence 61
- “Moreover, shipwreck, far from terrifying them, is their training. With the perils of the
sea they are not merely acquainted-they are familiarly acquainted”
The first sentence seems to be a clear statement that the Saxons used sails. The letters are
however edited before being published. Jones (1996) has argued that the writing is highly
poetic and he does therefore not find the text trustworthy (p. 92). Jones also states that is not
certain that Sidonius ever saw a Saxon vessel. To further support Sidonius´ ignorance of
Saxon vessels he refers to another text by Sidonius (Panegyric on avitus) where it is stated:
“The Aremorican region too expected the Saxon pirate, who deems it but sport to furrow the
British waters with hides, cleaving the blue sea in a stitched boat” (Sidonius, 1963, p. 151).
Jones interprets this to mean that Sidonius thought that the Saxons used sewn hide boats,
whilst other classical authors assign the hide boat to the British or Irish. Jones (1996)
therefore argues that Sidonius cannot be relied upon (p. 92). It is indeed doubtful whether
Sidonius´ remarks are first-hand descriptions. It may be that Sidonius met some Saxons in
Bordeaux, but this is no confirmation that he saw their vessels. Again the problem who is
exactly referred to is of importance. As discussed above the Saxons referred to may have been
primarily those no longer living in the North.
The second sentence has sometimes been interpreted as indicating that the Saxons did not use
sails. This is a doubtful inference. It may for example be that pirate attacks were mainly done
under oars or it was a way for Sidonius to emphasize the peculiarity that even an oarsman can
give orders which are obeyed. The latter may be supported by a similar statement by
Procopius that in the fleet of the angles there “were no supernumeraries … for all the men
rowed with their own hands.” (translation Dewing, p. 261). It does not exclude the use of the
sail.
The last sentence is often referred to as supporting the conclusions regarding Nydam B, i.e.
that the vessels of the Saxons were not suited for the open sea (Green, 1988, p. 54; Jones,
1996, p. 75). This sentence is taken out of context giving a different impression of what is
meant. Directly afterwards it is namely written:
for since a storm whenever it occurs lulls into security the object of their attack and
prevents the coming attack from being observed by victims, they gladly endure
dangers amid billows and jagged rocks in the hope of achieving a surprise. (translation
Anderson, 1965, p. 431)
It is therefore rather the situations they put themselves in rather than the capabilities of their
vessels. If indeed they made the choice to hide their vessel in storms and near jagged rocks it
rather indicates the reliance upon their vessels and their seamanship. Furthermore, it is not
known whether the Saxons used similar boats as Nydam B.
Again there are clear uncertainties and the use of the sentences is judged to be weak and so is
the relevance for the absence or presence of the sail in ´Scandinavia´.
Chapter 5. The arguments: 5.3 Historical evidence 62
5.3.4 Procopius
The sentence that is of importance here is: “And there were no supernumeraries in this fleet,
for all the men rowed with their own hands. Nor do these islanders have sails, as it happens,
but they always navigate by rowing alone.” (translation Dewing, p. 261).
The question is how reliable Procopius is as a source. Jones (1996) argues that he is reliable
since his description of vessels of the Vandals and Byzantines seem to be correct and he knew
that the Romans, Celtic people, the Vandals and the Goths used sails (p. 94). Jones further
argues that Procopius´ mention of the Angles being ignorant of horses and their habit of
fighting on foot, may give Procopius´ description further credibility.
Procopius´ knowledge about the vessels of the Vandals and Byzantines (which he witnessed
first-hand) hardly makes his statement on the Angles more reliable since it does not seem to
be a first-hand observation. His knowledge of the use of the sail regarding the Romans, Celts
and Goths could also have come from textual sources and the Goths were during the time of
writing situated near or at the Mediterranean Sea. That this strengthens Procopius´ credibility
regarding the vessel used on the Atlantic and North Sea is doubtful. It is not known and
perhaps rather unlikely, that Procopius ever saw an Anglian vessel, the same reason why
Jones discredits Sidonius´ statement regarding the Saxons.
Procopius other incorrect statement that the Angle had never seen a horse because these do
not live on Brittia neither strengthens his credibility. It may be that it is just an exaggeration to
point out that the Angles hardly used horses whilst fighting, but it may then also be that the
lack of sail is also just an exaggeration. As mentioned it may also be that the embassy had
reasons to present Angles as submissive to the Franks and may have portrayed them as
ignorant people. The fact that the statements of the absence of horses and sails are made
shortly after one another may support this.
Lebecq (1990), one of the proponents for the late use of the sail in north-west Europe states
that the alleged absence of the sail also regards the Varni (p. 85). This is however incorrect,
but this thought seems to also been taken over by Bill (2010, p. 28). From the context of the
sentence it is clear that it refers only to the fleet. Although Lebecq refers to Dewing, he gives
his own translation which is more in favour of his interpretation that the sentence also refers
to the Varni.
Procopius also tells a story concerning the ferrying of souls. Some have referred to this to
discredit Procopius´ account (Bengtsson, 2017, p. 48; Haywood, 1991, p. 70). However
Procopius clearly states his disbelief: “I shall presently explain, having many a time heard the
people there most earnestly describe it, though I have come to the conclusion that the tales
they tell are to be attributed to some power of dreams.” (translation Dewing, 1962, p. 267).
The probability of Procopius´ statement seem all in all weak. As for the relevance for the
thesis the statement refers to a group of Angles living in Britain in the 6th century, whether
this reflects the situation in Scandinavia is unsure and remains restricted to the Angles in
Britain. The relevance is therefore also rather weak.
Chapter 5. The arguments: 5.4 Assessment of the prevailing theory 63
5.4 Assessment of the prevailing theory
The sum of the arguments that have been used to support the theory that the sail was a late
adoption is weak. The vessel remains discussed, Hjortspring, the Nydam vessels and
Kvalsund II do not exclude the possibility of the sail. Neither do the Gotlandic picture stones
and the historical sources. The historical sources can neither be said to support an earlier use
of the sail in Scandinavia. The only evidence currently available that may speak against the
late adoption of the sail is the rock art, primarily dating to the Bronze Age. I have judged the
argument that it supports an earlier adoption as being ´good´. It is important to stress that this
is my own subjective interpretation of the rock art depictions. It is up to the reader to decide
whether to agree or not. The relevance of the premise depends on whether they depict local
vessels employing a sail or vessels from outside Scandinavia. If the depictions are accepted as
depicting sailing vessels, it is likely that this is the main point that can be criticised.
Nonetheless, none of the other arguments currently exclude the possibility of an earlier
adoption of the sail.
64
Chapter 6. Discussion
The constructional criteria as discussed in chapter five are often advocated as being a
necessity for a vessel if it is to be sailed. It is evident that this is not the case. The features are
mainly a necessity for oceanic voyages, but not for more sheltered coastal waters. The focus
has throughout the discourse been on the former. This becomes apparent from the description
by Brøgger and Shetelig (1950) of Hjortspring as a primitive vessel which would clearly have
been too frail to endure such strains and is too open and vulnerable in the open sea when
compared to the Gokstad and Tune ships (p. 57). The criteria do increase the sailing
capabilities of a vessel, but, as has been shown with the Tilia, the absence of these does not
exclude the possibility of it having been sailed. It neither excludes the possibility of it being
sailed 90 degrees to the wind and possibly even 60 degrees. The focus has especially been on
the windward performance of the vessels (Bengtsson, 2017, p. 44). It has been stated that
windward performance is also imporant for coastal areas for leaving or making a landing
place or for rounding a headland (McGrail, 1987, p. 220). This is however mainly of
importance when a vessel relies solely on sailing not when it is used in combination with
paddles or oars. When necessary one can then easily switch over to rowing or paddling. The
sail may have been no more than a helping device, a supplement to paddling or rowing. They
may have had a sailing practice very different to ours (Bengtsson, 2017, pp. 100, 119). This
may be indicated by the focus of the Scandinavian boatbuilders on having a vessel that can
both be sailed and rowed.
The method of setting up a list of criteria is a valid one. These criteria need to be made
explicit, so they may be criticised. As regards the criteria for a sailing vessels to cross the
ocean most of the criteria, such as high-freeboard and a deep and securely fastened rudder,
still hold. The need for a ´proper´ keel, i.e. not a keel plank but a beam with its widest part
placed vertically, preferably a deep protruding one, is questionable. The leeway resistance can
also be solved in other manners than only with a protruding keel. This may be done with the
shape of the hull, with deep rudders or other such devices. The amount of leeway also
depends on the type and size of sail. The sail may have been smaller, wider and not as high up
the mast as has been assumed which results in a lower centre of effort. For oceanic voyages a
vessel needs to have sufficient longitudinal and transverse support. A keel is an important
feature to provide longitudinal strength. A ´proper´ keel, i.e. a beam with its widest part
standing vertically, is stronger than with it laying horizontally. However making it thick
enough can compensate for this. Furthermore there are other methods of providing additional
longitudinal strength such as thicker strakes (a meginhufr for example) or stringers.
Concluding that a vessel could not cross the ocean because it had a keel plank cannot be so
easily stated. For the transverse strength beams would have been necessary.
Vessels are complex constructions and drawing conclusions on the basis of reconstruction
drawings is difficult and often inaccurate. It is mainly by actually building the vessel and
testing them in the water that it is possible to come closer to certainty. That is, if the
reconstructions are correct in the first place. Reconstructing a complete vessel from a few
fragments is difficult and often leads to several possibilities which may very well be incorrect
(Christensen, 1982, p. 24). There may have been constructional features which have not been
left to us either through preservation or through actions by people prior to the deposition.
Reconstructions are mainly based on the idea that no more should be reconstructed than what
can be inferred from the remains to avoid imaginative constructional features (Crumlin-
Pedersen, 2003b, p. 217). Yet this does not mean that these reconstructions are correct. This
Chapter 6. Discussion 65
has also been criticised by Bengtsson (2017, pp. 57–58). One should be aware that it is simply
a possibility. The strength of this possibility is mainly dependent on the state of preservation.
The lesser that remains the more careful one should be with simply accepting a reconstruction
as reflecting reality. It is, however, often the cased that once a reconstruction has been made
conclusions are drawn on the basis of these.
As thoroughly discussed the representativeness of the vessels has hardly been taken into
account. That a vessel is representative for the whole of ´Scandinavia´ has been too readily
accepted not taking into account the possibility of choice and thereby diversity. That they are
representative and that the Scandinavian boatbuilding tradition is conservative has been
supported by the evolutionistic development as advocated by Shetelig and Brøgger (1950).
This theory is based on vessels from different regions in Scandinavia and from different
periods, but have been placed in a single line of development. The basis of the theory itself
therefore rest on the idea that Scandinavia is one common region were technology develops
along the same line. This is an assumption. Hjortspring and the Nydam vessels are thought to
belong to Southern Scandinavia. Kvalsund II was found near the city of Ålesund in Norway
and Oseberg and Gokstad in Southern Norway. The flat keel may therefore just as likely have
been a specific feature for Danmark with its sandy beaches and the T-keel for Norway where
such beaches hardly exist. Although the vessels all belong to the clinker-built ´tradition´ it
does not mean that a development in one region also occurred in the other. There are
limitations to what we can infer from the archaeological record. If one wants to see the
vessels as representative for regions and periods other than those in which the vessels were
found and constructed it needs to be thoroughly argued for. Nydam B can for example not be
taken as representative for the vessels that were used for the crossing to Britain simply
because there are no other complete vessels found from the period.
Furthermore a vessel is built for an environment and for a function. Conscious choices were
made by the boatbuilders and those who ´ordered´ the vessel. Possibly Nydam B was
specifically built for the more protected waters to the east of Denmark. Nydam B does not
exclude the possibility that more seaworthy vessels were built on the west coast of Denmark
for the rougher waters of the North Sea. A shallow keel may have been preferred for regions
with shallow waters or when there is a need for often beaching a vessel. This may for example
be indicated by the shallow keel of Gredstedbro which show traces of heavy wear (Ejstrud &
Maarleveld, 2008, p. 63). That the vessels of Scandinavia prior to Gokstad were all-purpose
vessels has also been assumed thus excluding the possibility that they were constructed for a
certain function (Christensen, 1982, pp. 25–26). Hjortspring, the Nydam vessels and
Kvalsund II are all from a similar find context. It may therefore be a bias in the archaeological
material, i.e. the more common vessels have not been preserved. The few fragments that may
be from more common vessels are reconstructed on the basis of vessels that have been well
preserved which is obviously problematic for identifying other types of vessels. (see section
5.1.6).
The evolutionistic view has put the changes in the constructional features in direct association
with a change in the method of propulsion. The thought is however not that a new method of
propulsion led to a change in construction, but rather that a change in construction was made
in order to eventually adopt a new method of propulsion (Bengtsson, 2017, p. 56). Since it is
clear that none of the constructional criteria are necessary for the use of the sail it seem more
logical, in my opinion, that the sail was used on existing vessels which gradually developed
towards more advanced sailing vessels. Any theory of an earlier use of the sail needs as every
theory to be argued for and evidence provided. However, the absence of an earlier us of the
Chapter 6. Discussion 66
sail does not imply a late use of the sail. Thus, based on the current ship finds, any statements
that the sail was a late introduction, cannot be stated with any certainty.
As discussed, it may also be difficult to even recognise that a vessel has been sailed. What is
one supposed to look for? This is also applicable to the iconography. One should not look
with preconceptions. The only possible evidence for an earlier use of the sail are currently the
depictions on the rock art, the hvirring razor and the vendsyssel carving. Whether these
indeed depict sails is subjective, but the possibility cannot be excluded. That the first evidence
of sailing vessels is from a later period should not be used as an argument that they do not
represent sails. Such circular reasoning should be avoided.
The historical sources neither provide a conclusive answer. Apart from their trustworthiness it
is also important to discuss to what extent their statements can be generalised. They mainly
discus a certain people. The statement can therefore not be taken to also include the rest of
Scandinavia. As for the ´Suiones´ there seems to be an agreement that this refers to a people
in current Sweden. As for the ´Angles´ or ´Saxons´ it is the question whether the authors
specifically meant those people or that it had become a more general term for pirates. Those
texts that mention these people date to the migration period. It is therefore unknown whether
the authors refer to people still living in the North or those that already have established
themselves in Britain. To generalise any statement regarding these people to also include the
rest of ´Scandinavia´ should be argued for. ´Scandinavia´ is a modern construct but within
archaeology it has all too often been seen as one cultural region that was also present in the
past. This also seems to be the case within the current discourse. Furthermore, when the texts
refer to the Saxons it is mainly in the context of their piratic activities. The sentences may
therefore only refer to the vessels used for such activities.
The current method that is being employed in the mainstream discourse seems to be that of
falsification, i.e. an earlier date is not accepted until evidence is published that unequivocally
proves an earlier use of the sail. Popper´s method of falsification was however based on the
possibility of running endless tests. This is something that is not possible with the
archaeological record. The use of such a method within archaeology is therefore, to my liking,
to simplistic and has a stagnating effect on any discussion. Due to only a few finds being left
to us from the past, the representativeness of these should be taken into account before stating
any generalising theories. As was discussed in chapter two, an earlier use of the sail can never
be disproved. Ergo a late usage of the sail can never be proven. This is clearly problematic
and the reader may wonder; if it indeed was a late adoption how may it ever be known for
certain that it was? Inductive reasoning can of course never be certain, it may however have a
high probability. The probability increases proportional to the increased representativeness of
the archaeological record. The question is of course when to know that it is representative.
Currently only a few relatively complete vessels have come to light. That these vessels are
representative for those in Scandinavia for their respective period can be said to not have been
well argued for. Neither is it certain that these vessels were never sailed, there may even be
indications that they were such as the cleats of the Hjortspring (section 4.1.1.5). The
prevailing theory rests on an unstable basis.
67
Chapter 7. Conclusion
The main aim of this thesis is to determine the strength of the theory that the sail was adopted
late in Scandinavia. The focus is on the arguments that are directly based on the evidence that
can inform us on the use of the sail in Scandinavia. Arguments based on other theories were
excluded. The evidence is divided in three categories; vessel remains, iconography and textual
evidence. Each of these types of evidence have their own difficulties as evidence.
The absence of evidence has been taken as an indication that the sail was not employed in
Scandinavia until very late in the Iron Age. If not specifically stated it is often implied that
vessels prior to Oseberg where not propelled by a sail. This is mainly based on Hjortspring,
Nydam B and Kvalsund II, which according to the mainstream discourse do not have any
evidence of having been sailed. Identifying whether a vessel was used for sailing may be very
difficult. More difficult than has been acknowledged. Nothing more than wear may be
present. Even when evidence is present it may be difficult to identify it as such. Possibly the
cleats at the stems of Hjortspring were used for stays and in the case of Kvalsund II a mast
may even have been found, but it has never received any attention.
What was looked for? It seem that the later Oseberg and Gokstad and the depictions on the
Gotlandic picture stones have led to several assumptions. Which are; that Scandinavian
sailing vessels always used a large square sail, that they had one mast placed midship in a
keelson with a mast partner and that the constructional features present in these vessels are
perquisites for a sailing vessel. The question that has been discussed until now is not so much
when the first sail was used in Scandinavia, but rather when the people in Scandinavia had
sailing vessels which could be used for oceanic voyages. These are two different topics which
have been mixed up in the discourse. The focus on oceanic voyages seems to stem from the
focus on the Migration- and Viking-period. Quite a few extra criteria for a vessel are then of
relevance such as, a deep protruding keel against the leeway which is necessary if one wants
to stay on course for a long period if no accurate control points are available, sufficient
longitudinal and transverse strength (strong keel, beams, etc.) to cope with the forces of the
sea and a stable hull (a hull with a wide or wine glass shaped cross-section rather than a round
one). These criteria should however not be seen as criteria for a vessel to sail, which has been
clearly demonstrated with the sailing trials of the Tilia, a reconstruction of the Hjortspring.
The representativeness of the vessels have not been taken into account in the discourse. The
vessels have been too readily accepted as representative for all the vessel for an entire period
in Scandinavia. Simply because there are only a few finds of nearly complete vessels in
Scandinavia does not mean that their representativeness should not be discussed. The
Hjortspring, Nydam B and Kvalsund II were all found in the same context which seem to
indicated that these functioned as warships/troop transporters. These vessels are likely built
specifically for this function and their environment. The use of a keel plank rather than a deep
protruding keel may therefore have been due to shallow waters and due to a need for easily
beaching a vessel. The possibility of regional diversity and choices by the boatbuilders have
through the evolutionistic development of the ´Scandinavian´ vessels been disregarded.
Nydam B does not exclude the use of more seaworthy vessels or the use of other types of
keels elsewhere during the 4th century (as possibly Nydam C indicates). They neither exclude
the possibility of other vessel types or the use of the sail on other vessels even if it is
determined that the above vessels were not sailed.
Chapter 7. Conclusion 68
If it is to be accepted that these are representative it needs to be further argued for. On the
basis of the current ship remains and arguments it is not possible to exclude the possibility
that sails were used in Scandinavia prior to Oseberg (820 AD).
The iconography that has mainly been used as evidence in the discourse are the earliest
depictions of sailing vessels on the Gotlandic picture stones. Their date is uncertain but seems
to be either from the 6th or 7th century. These are seen as the first evidence of the sail in
Scandinavia in the mainstream discourse. Because there are also Gotlandic picture stones
from the 5th and 6th centuries depicting rowing vessels similar in shape as Nydam B and Sutton
Hoo II these stones are thought to depict reality. Thus the shift from rowing to sailing vessels
on the stones has been interpreted as an actual change in propulsion in Scandinavia. It is
questionable whether such an inference can be drawn. These stone seem to be of a religious
and mythological character. A special type of vessel may therefore be represented rather than
a commonly used vessel. Furthermore, that the Gotlandic picture stones also reflects the
situation in the rest of Scandinavia is an assumption which should be argued for not merely
accepted.
There are possible depictions of sailing vessel in the rock art of Scandinavia. These have
however hardly been mentioned in the mainstream discourse except when stating that they do
not resemble sailing vessels. The figures above the vessels are more often seen as depicting
other symbols thereby excluding the possibility that these depict mast and sail. Since it has
been shown that the constructional features of the Viking ships do not have to be present in a
vessel for it to sail, the possibility cannot be excluded that the rock art includes depictions of
sailing vessels. Identifying figures in rock art remains is to a large extent interpretative. It is
up to the reader to decide whether these depict sailing vessels or not. One should be open to
the idea that different sizes and shapes of sails could have been used, not only square sails.
These same issues also apply to the Hvirring razor and the vendsyssel vessel.
Apart from these depictions there is also the Karlby pebble and the Weser bone. Due to the
unknown date of the Karlby pebble it cannot be used as evidence for an earlier adoption of the
sail. The Weser bone from the 4th or 5th century does imply that the people along the Weser
were aware of the usage of the sail, but it cannot be concluded on the basis of this that the sail
was used in ´Scandinavia´.
Lastly there is the textual evidence. In Germania, from 98 AD, Tacitus describes the states of
the Suiones. He states that they do not use sails, either referring to all of the vessels of the
Suiones or only their warships depending on the translation. Although there seems no reason
for Tacitus to alter the truth, the information he received himself is likely unreliable. Tacitus
mentions that their vessels differs from the norm which seems to indicate that their vessels
differ from the vessels of the other Germanic people. Even if the information is correct it can
hardly be generalised to include the rest of Scandinavia.
The text by Gildas from the first half of the sixth century is highly literate and due to an
uncertainty in translation of the relevant words its use as evidence for an earlier use of the sail
is doubtful. Several sentences of the text by Sidonius from 480 AD have been interpreted to
either support or disprove the use of the sail. None of these sentences seem to hold as
evidence for either. Both these texts regard the ´Saxons`. It is uncertain to who this term
exactly refers to. Whether this refers to the already migrated Saxons or whether it had already
become a term synonymous to pirates. Although the area of the Saxons is outside
Chapter 7. Conclusion 69
Scandinavia, following the modern borders, the statements have been generalised to include
Scandinavia. This is a questionable generalisation.
A text by Procopius concerns the vessel of the Angles living on Brittia. Procopius state that
they do not have sails but only row. The information seems to have come from an Frankish
embassy in Constantinople. The aim of this embassy was to assert the rule of the Franks over
the Angles. Although Angles were included in the embassy it is questionable whether the
truth was not distorted. Shortly after mentioning the absence of the sail it is said that no horses
lived on Brittia and the overall clumsiness of the Angles as regards the horses is described. It
may therefore well have been that the Franks aimed to portray the Angles as being less
advanced than the Franks. It is therefore an unreliable source. The textual evidence provides
therefore no conclusive answer.
The theory that the sail was a late adoption in Scandinavia rests therefore on an unsolid basis.
It is relies on several assumptions. The main assumption being that all the people in
Scandinavia were alike, so that evidence from one region is generalised to also include the
rest of Scandinavia. This is visible in the arguments of the textual evidence, the iconography
and the vessel remains. ´Scandinavia` is a modern construct. The strength of such
generalisations depends on the strength of the theory that Scandinavia was a common cultural
region in prehistory. This should be underpinned with evidence and arguments. Even so, the
possibility of regional diversity cannot be excluded.
The primary method used in the discourse is that of falsification, i.e. an earlier date is not
accepted until evidence is found that unequivocally indicates an earlier date. Although this
may be a ´scientific´ method, it has a stagnating effect on the discourse. Archaeology is a
discipline of interpretation with has to draw conclusion from only a few fragments left to us.
Although there may not be evidence from an earlier date, it is never redundant to discuss the
uncertainties of a theory.
70
Glossary
The following definitions are taken over
from Bengtsson, 2017, Ditta et al., 2014
and Steffy, 2012.
Athwartships – Across the ship from side
to side; perpendicular to the keel.
Ballast – Heavy material, such as iron,
lead or stone, placed low in the hold to
lower the centre of gravity and improve
stability.
Beam – A timber mounted athwartships to
support decks and provide lateral strength.
Bow – The forward part of a hull,
specifically, from the point where the sides
curve inward to the stem.
Centre of effort – often referred to as the
´centre of pressure´ i.e. the point at which
the combined aerodynamic force of the
wind acts on a sail
Centre of gravity – affects a boat stability.
The higher up the centre of effort is on the
sail (i.e. the higher the sail is), the less
stable the boat will be. On a small and
narrow boat, a low centre of gravity is
preferred. This can be adjusted by
members of the crew sitting to windward,
counteracting the effects of the wind or, as
in Oceania, by widening the boat with the
use of e.g. outriggers.
Chock – An angular block or wedge use to
fill out areas between timbers or to
separate them.
Draft [Draught] – The depth to which a
hull is immersed; also, a drawing or plan.
Frame – Transvers timber to which the
planks are fastened.
Keel – the main longitudinal timber of
most hulls, upon which the frames,
deadwoods, and ends of the hull were
mounted; the backbone of the hull.
Keel plank – A central hull plank that was
substantially thicker than the rest of the
bottom planking and whose breadth was at
least twice as great as its thickness.
Keelson – An internal longitudinal timber
or line of timbers, mounted atop the frames
along the centreline of the keel, that
provided additional longitudinal strength to
the bottom of the hull. [note author: In the
early Viking vessels the keelson did not
run across the entire length of the vessel
but was fitted over several frames midships
and mainly functioned as a timber for
holding the mast in place and to transfer
the forces from the mast to the frames.]
Leeway – The sideways drift of a vessel
when sailing with the wind abeam.
Mast partner – The timbers surrounding
the deck openings for masts, pumps, bitts,
and capstans; their primary purpose was to
strengthen the deck around the opening
and counteract strain. Partners were also
used on occasion to steady masts on
undecked vessels.
Mast step – A mortise cut into the top of a
keelson or large floor timber, or a mortised
wooden block or assembly of blocks
mounted on the floor timbers or keelson,
into which the tenoned heel of a mast was
seated.
Midship [Midships] – A contraction of
amidships and consequently, in a general
sense, it refers to the middle of the ship. In
construction, however, it is often used as
an adjective referring to the broadest part
of the hull, wherever it may be.
Midship frame – The broadest frame in
the hull; the frame representing the
midship shape on the body plan.
Chapter 7. Conclusion 71
Rabbet – A groove or cut made in a piece
of timber in such a way that the edges of
another piece could be fit into it to make a
tight joint. Generally, the term refers to the
grooves cut into the sides of the keel, stem,
and sternpost, into which the garboards
and hooding ends of the outer planking
were seated.
Rove – A small metal washer, used in
clinker-built hulls, over which nail or rivet
ends are flattened to lock the fastening.
Rudder – A timber, or assembly of
timbers, that could be rotated about an axis
to control the direction of a vessel
underway.
Scarf – An overlapping joint used to
connect two timbers or planks without
increasing their dimensions.
Shroud – A rope or wire support used to
steady a mast to the side of a hull.
Stanchion – An upright supporting post,
including undecorated supports for deck
beams and bulkheads.
Steering oar – An oar to steer a small
vessels, either from the side or the stern.
Stem – A vertical or upward curving
timber or assembly of timbers, scarfed to
the keel or central plank at its lower end.
Stern – The after end of a vessel.
Strake – A continuous line of planks,
running from bow to stern.
Thwart – A transverse plank in a boat or
galley; used to seat rowers, support masts
or provide lateral stiffness.
Trussing rope – A rope that runs from
bow to stern of a ship or boat and which
can be tightened in order to provide extra
longitudinal strength.
72
Bibliography
Åkerlund, H. (1963). Nydamskeppen: en studie i tidig skandinavisk skeppsbyggnadskonst.
Göteborg: Sjöfartsmuseet i Göteborg.
Arbman, H. (1940). Birka: Untersuchungen und Studien. Die Gräber: Tafeln. (Vol. 1).
Stockholm: Almqvist & Wiksells Boktrykeri-Aktiebolag.
Audi, R. (Ed.). (1999). The Cambridge dictionary of philosophy (2nd ed). Cambridge ; New
York: Cambridge University Press.
Bengtsson, B. (2017). Sailing rock art boats: a reassessment of seafaring abilities in Bronze
Age Scandinavia and the introduction of the sail in the North. Oxford, UK: BAR
Publishing.
Bengtsson, B., & Bengtsson, B. (2007). Seglade man i skandinavien under bronsåldern? - om
experimentella seglingsförsök i ‘bronsåldersliknande’ farkoster. Marinarkeologisk
Tidskrift, (2–3), 14–17.
Bengtsson, B., & Bengtsson, B. (2011). Sailing Rock Art Boats. Journal of Maritime
Archaeology, 6(1), 37–73. https://doi.org/10.1007/s11457-011-9077-2
Bill, J. (1997). Ships and Seamanship. In P. H. Sawyer (Ed.), The Oxford illustrated history of
the Vikings (pp. 182–201). Oxford: Oxford University Press.
Bill, J. (2010). Viking Age ships and seafaring in the West. In I. S. Klæsøe (Ed.), Viking trade
and settlement in continental Western Europe (pp. 19–42). Copenhagen: Museum
Tusculanum Press, University of Copenhagen.
Brøgger, A. W., & Shetelig, H. (1950). Vikingeskipene: deres forgjengere og etterfølgere.
Oslo: Dreyers forlag.
Brøndsted, J. (1940). Danmarks oldtid : III : jernalderen. Kbh.: Gyldendal.
Bugge, A. (1923). Den norske sjøfarts historie : fra de ældste tider til vore dage. I bind. Oslo.
Chapter 7. Conclusion 73
Christensen, A. E. (1968). Boats of the North : a history of boatbuilding in Norway. Oslo.
Christensen, A. E. (1982). Viking Age Ships and Shipbuilding. Norwegian Archaeological
Review, 15, 19–28.
Christensen, A. E. (1984). Uten båt – ingen besetting. In S. 1943- Indrelid & S. U. 1938-
Larsen (Eds.), Sunnmøres forhistorie: fra de første fotefar (pp. 128-137). Ålesund:
Sunnmørspostens forlag.
Christensen, A. E. (1996). Proto-Viking, Viking and Norse Craft. In R. Gardiner (Ed.), The
earliest ships: the evolution of boats into ships (1st publ, pp. 72–88). London:
Conway Maritime Press.
Coll, A. L. (1902). Fra Helleristningernes Omraade. Foreningen Til Norske
Fortidsminnesmerkers Bevaring. Aarsberetning for 1901., 3–59.
Coll, A. L. (1903). Fra Helleristningernes Omraade (andet Stykke). Foreningen Til Norske
Fortidsminnesmerkers Bevaring. Aarsberetning for 1902., 106–140.
Crumlin-Pedersen, O. (1965). Cog-Kogge-Kaag. Handels- Og Søfartsmuseet På Kronborg:
Årbog 1965, 81–144.
Crumlin-Pedersen, O. (1968). The Gredstebro Ship: The remains of a Late Iron Age Vessel
Found in 1945 in South Jutland. Acta Archaeologica, 262–267.
Crumlin-Pedersen, O. (1990). The boats and ships of the Angles and Jutes. In McGrail (Ed.),
Maritime Celts, Frisians, and Saxons: papers presented to a conference at Oxford
in November 1988 (pp. 98–116). London: Council for British Archaeology.
Crumlin-Pedersen, O. (2003a). Initial analysis and reconstruction of the boat. In A. Trakadas
& O. Crumlin‐Pedersen (Eds.), Hjortspring: A Pre-Roman Iron-Age Warship in
Context (pp. 23–36). Roskilde: Viking Ship Museum Roskilde.
Crumlin-Pedersen, O. (2003b). The Hjortspring boat in a ship-archaeological context. In A.
Trakadas & O. Crumlin‐Pedersen (Eds.), Hjortspring: A Pre-Roman Iron-Age
Warship in Context (pp. 209–233). Roskilde: Viking Ship Museum Roskilde.
Chapter 7. Conclusion 74
Crumlin-Pedersen, O. (2010). Archaeology and the sea in Scandinavia and Britain: a
personal account. Roskilde: Viking Ship Museum Roskilde.
Dahlgren, P. (1932). Primitiva skepp : vad hällristningarna vittna om bronsålderns sjökultur.
Stockholm: Natur och Kultur.
Ditta, M., Firth, A., Nayling, N., Ní Chiobháin, D., Thomsen, C., & Wagstaffe, C. (2014). The
Gresham Ship project: a 16th-century merchantman wrecked in the Princes
Channel, Thames Estuary. Volume 1: Excavation and hull studies. (J. Auer & Th.
J. Maarleveld, Eds.). Oxford, England: Archaeopress.
Ejstrud, B., & Maarleveld, Th. J. (2008). The migration period, Southern Denmark and the
North Sea: a workbook in relationship to the Gredstedbro find. Esbjerg: Maritime
Archaeology Programme, University of Southern Denmark.
Ellmers, D. (1972). Frühmittelalterliche Handelsschiffahrt in Mittel- und Nordeuropa.
Neumünster: Wachholtz.
Engelhardt, C. (1865). Nydam mosefund: 1859-1863. Copenhagen: Thieles bogtrykkeri.
Farrell, A. W. (1979). Mast and Sail in Scandinavia in the Bronze Age? The Mariner’s
Mirror, 65(1), 83.
Fenger, N. P. (2003). Geometrical properties. In A. Trakadas & O. Crumlin‐Pedersen (Eds.),
Hjortspring: A Pre-Roman Iron-Age Warship in Context (pp. 89–92). Roskilde:
Viking Ship Museum Roskilde.
Fenger, N. P., & Hocker, F. M. (2003). Stability. In A. Trakadas & O. Crumlin‐Pedersen
(Eds.), Hjortspring: A Pre-Roman Iron-Age Warship in Context (pp. 92–94).
Roskilde: Viking Ship Museum Roskilde.
Flierman, R. (2017). Saxon identities, AD 150-900. London ; New York: Bloomsbury
Academic, an imprint of Bloomsbury Publishing Plc.
Gildas, & Winterbottom, M. (1978). The ruin of Britain and other works. London: Phillimore.
Giles, J. A., & Gildas. (1841). The Works of Gildas and Nennius. London: James Bohn.
Chapter 7. Conclusion 75
Glob, P. V. (1969). Helleristninger i Danmark (Vol. 7). København: Nordisk forlag.
Göttlicher, A. (1989). Roman ship on Runic bone. International Journal of Nautical
Archaeology, 18(1), 73.
Green, C. (1988). Sutton Hoo: the excavation of a royal ship-burial (2. rev. ed). London:
Merlin Pr.
Greenhill, B. (1976). Archaeology of the boat: a new introductory study. London: A. and C.
Black.
Halldin, G. (1952). Vad för slags farkoster återge de sydskandinaviska hällristningarnas
skeppsbilder? Sjöhistorisk Årbok 1952, 11–90.
Hawkes, C. (1954). Archeological Theory and Method: Some Suggestions from the Old
World. American Anthropologist, 56(2), 155–168.
https://doi.org/10.1525/aa.1954.56.2.02a00660
Haywood, J. (1991). Dark age naval power: a re-assessment of Frankish and Anglo-Saxon
seafaring activity. London ; New York: Routledge.
Haywood, J. (2006). Dark age naval power: a reassessment of Frankish and Anglo-Saxon
seafaring activity (Rev. and extended ed., reprinted). Norfolk: Anglo-Saxon Books.
Hocker, F. M. (2003). Documentation of the form and structure of the hull. In A. Trakadas &
O. Crumlin‐Pedersen (Eds.), Hjortspring: A Pre-Roman Iron-Age Warship in
Context (pp. 209–233). Roskilde: Viking Ship Museum Roskilde.
Holmboe, J. (1929). Fundforholdene botanisk undersøkt. In H. Shetelig & F. Johannessen,
Kvalsunfundet: og andre norske myrfund af fartøier (pp. 17–21). Bergen.
Humbla, P., & von Post, L. (1937). Galtabäcksbåten och tidigt båtbyggeri i Norden. Kungliga
Vetenskaps- Och Vitterhets-Samhälles Handlingar. Femte Följden. Seria A., 6(1).
Imer, L. M. (2004). Gotlandske billedsten - dateringen af Lindqvist’s gruppe C og D.
Aarbøger for Nordisk Oldtidskundskab Og Historie 2001, 47-111.
Chapter 7. Conclusion 76
Jessen, K. (1929). Nelden (Urtica dioeca L.) i Kvalsund-Fundet. In H. Shetelig & F.
Johannessen, Kvalsunfundet: og andre norske myrfund af fartøier (pp. 17–21).
Bergen.
Johnstone, P. (1980). The seacraft of prehistory. London: Routledge & Kegan Paul.
Jones, M. E. (1996). The end of Roman Britain. Ithaca, N.Y.: Cornell Univerity Press.
Kastholm, O. T. (2007). Viking Age Iconography and the Square Sail. Maritime Archaeology
Newsletter from Denmark, 22, 8-12.
Kaul, F. (2003a). The Hjortsping boat and ship iconography of the Bronze Age and Early Pre-
Roman Iron Age. In A. Trakadas & O. Crumlin-Pedersen (Eds.), Hjortspring: A
Pre-Roman Iron-Age Warship in Context (pp. 187–207). Roskilde: Viking Ship
Museum Roskilde.
Kaul, F. (2003b). The Hjortsping find. In A. Trakadas & O. Crumlin-Pedersen (Eds.),
Hjortspring: A Pre-Roman Iron-Age Warship in Context (pp. 141–185). Roskilde:
Viking Ship Museum Roskilde.
Lebecq, S. (1990). On the use of the word ´Frisian´ in the 6th-10th centuries written sources:
some interpretations. In Sean McGrail (Ed.), Maritime Celts, Frisians, and Saxons:
papers presented to a conference at Oxford in November 1988 (pp. 85–90).
London: Council for British Archaeology.
Lindqvist, S. (1933). Gotlands bildstenar. Rig, 97–117.
Lindqvist, S. (1941). Gotlands Bildsteine (Vol. 1). Stockholm: Wahlström och Widstrand.
Lindqvist, S. (1942). Gotlands Bildsteine (Vol. 2). Stockholm: Wahlström och Widstrand.
Maarleveld, T. J. (2010). Fish and ‘Chips of Knowledge’: Some Thoughts on the Biases of
the Archaeological record. In T. Bekker-Nielsen & D. Bernal Casasola (Eds.),
Ancient nets and fishing gear: proceedings of the International Workshop on ‘Nets
and Fishing Gear in Classical Antiquity: A First Approach’, Cádiz, November 15-
Chapter 7. Conclusion 77
17, 2007 (pp. 257–271). Cádiz : Aarhus, Denmark: Servicio de Publicaciones de la
Universidad de Cádiz ; Aarhus University Press.
Marstrander, S. (1963). Østfolds jordbruksristninger: Skjeberg. Oslo: Universitetsforlaget.
Marstrander, S. (1976). Building a hide boat. An archaeological experiment. International
Journal of Nautical Archaeology, 5(1), 13–22. https://doi.org/10.1111/j.1095-
9270.1976.tb00934.x
Martens, J. (2001). A Wooden Shield Boss from Kvärlöv, Scania. Some Remarks on the
Weaponry of the Early Pre-Roman Iron Age in Northern Europe and the Origins of
the Hjortspring Warriors. In M. Meyer & A. Leube (Eds.), ‘ --Trans Albim
fluvium’: Forschungen zur vorrömischen kaiserzeitlichen und mittelalterlichen
Archäologie: Festschrift für Achim Leube zum 65. Geburtstag. Rahden, Westf:
Leidorf.
McGrail, S. (1987). Ancient boats in N.W. Europe: The archaeology of water transport to AD
1500. London: Longman.
McGrail, S. (2001). Boats of the world: from the Stone Age to Medieval times. Oxford:
Oxford University Press.
Montelius, O. (1906). Kulturgeschichte Schwedens: von den ältesten Zeiten bis zum elften
Jahrhundert nach Christus. Leipzig: E. A. Seemann.
Moortel, A. van de. (1991). A cog-like vessel from the Netherlands. Lelystad: RWS, FL.
Müller, S. (1897). Nordische Altertumskunde : Nach Funden und Denkmälern aus Dänemark
und Schleswig. Strassburg.
Nancke-Krogh, S. (1987). Båtar och skepp: Bilder och verklighet. Fynd, 87(1), 2–13.
Nerman, B. (n.d.). Gravfynden på Gotland under tiden 550-800 e. K. Antikvarisk Tidskrift För
Sverige, 22(4).
Nicolaysen, N. 1817-1911. (1882). Langskibet fra Gokstad ved Sandefjord. Kristiania:
Cammermeyer.
Chapter 7. Conclusion 78
Nylén, E., & Lamm, J. P. (1988). Stones, ships and symbols. Stockholm: Gidlunds Bokförlag.
Nylén, E., & Lamm, J. P. (2003). Bildstenar (3. rev. uppl.). Stockholm: Gidlunds.
Pieper, P. (1989). Die Weser-Runenknochen: neue Untersuchungen zur Problematik: Original
oder Fälschung. Oldenburg: Isensee.
Popper, K. R. (1968). The logic of scientific discovery (2.). London: Harper & Row.
Procopius, & Dewing, H. B. (1962). Procopius: History of the Wars (Vol. 5). Cambridge,
Massachusetts: Harvard University Press.
Randsborg, K. (1995). Hjortspring: Warfare and Sacrifice in Early Europe. Aarhus: Aarhus
University Press.
Rausing, G. (1984). Prehistoric boats and ships of northwestern Europe: some reflections.
Malmö, Sweden: CWK Gleerup.
Reinders, R. (1985). Cog finds from the Ijsselmeerpolders. Lelystad.
Reitan, M. (2009). Argumentasjonsteori. Trondheim: Tapir Akademisk Forlag.
Rieck, F. R., & Crumlin-Pedersen, O. (1988). Både fra Danmarks oldtid. Roskilde:
Vikingeskibshallen.
Rosenberg, G. (1937). Hjortspringfundet (Vol. 3:1). København: Gyldendal.
Salin, B. (1904). Die altgermanische Thierornamentik : Typologische Studie über
germanische Metallgegenstände aus dem IV bis IX Jahrhundert, nebst einer Studie
über irische Ornamentik. Stockholm.
Sawyer, P. H. (1962). The age of the Vikings. London: Edward Arnold.
Schovsbo, P. O. (1980). Mast and Sail in Scandinavia in the Bronze Age? The Mariner’s
Mirror, 66(1), 15–16.
Chapter 7. Conclusion 79
Shetelig, H., & Johannessen, F. (1929). Kvalsunfundet: og andre norske myrfund af fartøier.
Bergen.
Shetelig, H., & Johannessen, F. (1930). Das Nydamschiff. Acta Archaeologica I, 1–31.
Sidonius, A. & Anderson, W. B. (1963). Sidonius: Poems and Letters. (Vol. 1). Cambridge,
Massachusetts: Harvard University Press.
Sidonius, A., & Anderson, W. B. (1965). Sidonius: Poems and Letters (Vol. 2). Cambridge,
Massachusetts: Harvard University Press.
Sidonius, A., & Dalton, O. M. (1915). The Letters of Sidonius (Vol. 2). Oxford: Oxford
University Press.
Sognnes, K. (2001). Prehistoric Imagery and Landscapes. Rock art in Stjørdal, Trøndelag,
Norway.
Solberg, B. (2003). Jernalderen i Norge: ca. 500 f. Kr.-1030 e. Kr. (2nd ed). Oslo: Cappelens
forlag.
Steffy, J. R. (2012). Wooden ship building and the interpretations of shipwrecks (1. ed., 4.
print). College Station: Texas A & M University Press.
Stölting, S. (1996). Überlegungen zu Felsbildschiffen mit Mast oder Segel. Adoranten, 31–35.
Stölting, S. (1999). Überlegungen zum Bau und zur Entwicklung bronzezeitlicher Schiffe.
Deutches Schiffahrtsarchiv, 22, 265–284.
Tacitus, Mattingly, H. B., & Rives, J. B. (2010). Agricola ; Germania. London: Penguin.
Tacitus, P. C., & Peterson, W. (1914). Dialogus. Agricola. Germania. London: Heinemann.
Trakadas, A., & Crumlin‐Pedersen, O. (Eds.). (2003). Hjortspring: A Pre-Roman Iron-Age
Warship in Context. Roskilde: Viking Ship Museum Roskilde.
Varenius, B. (1992). Det nordiska skeppet: Teknologi och samhällsstrategi i vikingatid och
medeltid. Stockholm.
Chapter 7. Conclusion 80
Vinner, M. (2003a). Sea trials. In A. Trakadas & O. Crumlin‐Pedersen (Eds.), Hjortspring: A
Pre-Roman Iron-Age Warship in Context (pp. 103–118). Roskilde: Viking Ship
Museum Roskilde.
Vinner, M. (2003b). Testing. In A. Trakadas & O. Crumlin‐Pedersen (Eds.), Hjortspring: A
Pre-Roman Iron-Age Warship in Context (pp. 134–136). Roskilde: Viking Ship
Museum Roskilde.
Vogel, W. (1915). Geschichte der deutschen Seeschiffahrt. Berlin.
von der Porten, E. (1978). Mast and Sail in Scandinavia in the Bronze Age? The Mariner’s
Mirror, 64(3), 284.
Westerdahl, C. (1995). Society and Sail. On symbols as specific social values and ships as
catalysts of social units. In O. Crumlin-Pedersen & B. M. Thye (Eds.), The ship as
symbol in prehistoric and medieval Scandinavia : papers from an international
seminar at the Danish National Museum, Copenhagen 5th-7th May 1994 (pp. 41–
50). Copenhagen: National Museum of Denmark, Department of Archaeology and
Early History.
Westerdahl, C. (2008). Boats Apart. Building and Equipping an Iron‐Age and Early‐Medieval
Ship in Northern Europe. International Journal of Nautical Archaeology, 37(1),
17–31. https://doi.org/10.1111/j.1095-9270.2007.00170.x
Westerdahl, C. (2015). The Viking Ship in Your Mind. Some comments on its cognitive
roles., 33–70.
Worsaae, J. J. A. (1859). Nordiske Oldsager i Det Kongelige Museum i Kjøbenhavn (Forøget
Udgave af ‘Afbildninger fra det Kgl. Museum for nord. Oldsager i Kjøbenhavn
1854’). Kjøbenhavn: Kittendorff & Aagaards Forlag.
