Content uploaded by Elena Karagjosova
Author content
All content in this area was uploaded by Elena Karagjosova on Mar 26, 2020
Content may be subject to copyright.
International Review of Pragmatics 3 (2011) 33–58
© Koninklijke Brill NV, Leiden, 2011 DOI 10.1163/187731011X561018
brill.nl/irp
Discourse Particles, Discourse Relations
and Information Structure:
e Case of Nämlich *
Elena Karagjosova
University of Stuttgart, Germany
elena.karagjosova@ling.uni-stuttgart.de
Abstract
e paper presents an analysis of the meaning and discourse eff ects of the German discourse
particle nämlich that unifi es its diff erent readings and explains its distributional properties.
I suggest that nämlich is most adequately analysed in terms of it indicating a specifi cational rela-
tion between its host and the preceding sentence, which in a question-based framework can be
implemented as indicating an answer to a “specifying question”, a discourse question requiring
an answer that provides a more detailed description of some aspect of the preceding utterance.
e analysis represents a refi nement and extension of the question-based analysis of nämlich
developed in Onea and Volodina (2009) where nämlich is analysed in terms of indicating that its
host is a short answer to an implicit constituent question or a Why -discourse question. e
approach I provide suggests solutions to several puzzles related to the distributional properties
of nämlich .
Keywords
specifi cation , causality , event coreference , discourse questions , discourse particle
1. Introduction
A typical and puzzling property of discourse particles is that depending
on their syntactic position they display diff erent semantic and pragmatic
properties. e particle nämlich is a case in point. It has been argued that
* I would like to thank the audience of the International Conference on Reference and
Discourse Structure, Stuttgart 2010, as well as Edgar Onea and Ken Turner and two anonymous
reviewers for comments and helpful discussion. In addition, I would like to thank the organizers
of the International Conference on Reference and Discourse Structure, Klaus von Heusinger
and Sofi ana Chiriacescu, for additional comments, help and support. Of course, all shortcom-
ings are my own. is research has been supported by the German Science Foundation (Project
C4 “Discourse particles. Incremental interpretation in context” of the SFB 732 “Incremental
Specifi cation in Context“) which I gratefully acknowledge.
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
34 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
depending on the particular syntactic environment in which it is used, näm-
lich may receive one of the following readings: a “specifying” reading, a “causal”
reading, and a “topic shift” reading.
In the following, I present an analysis of the semantic and pragmatic prop-
erties of nämlich which unifi es its alleged diff erent readings and provides a
natural explanation of the topic shift eff ect it displays in certain environments.
e account presented here is informed by considerations concerning the
interplay between discourse structure and discourse coherence on the one
hand and information structure on the other and suggests that the nature of
discourse particles can only be adequately captured from such an integrated
research perspective.
e approach I present is based on the analysis of nämlich suggested in
Onea and Volodina ( 2009 ), which I introduce in Section 2 and take as a point
of departure for the more elaborate and empirically adequate account that
I present in sections 3 and 4. Section 5 provides conclusions and some open
questions.
2. Onea and Volodina ( 2009 )
In a recent paper, Onea and Volodina ( 2009 ) provide an analysis of the
German discourse particle nämlich that diff ers from earlier analyses of the
particle (Hartmann, 1977 ; Granito, 1984 ; Vinckel, 2008 ) in its considerable
degree of detail and complexity. It is moreover cast in a question based model
of discourse, such as the ones developed by Roberts ( 1996 ) and Beaver and
Clark ( 2008 ) and which has been successfully used as a framework for analysing
discourse connectors such as but and and in several languages (Umbach, 2005 ;
Jasinskaja und Zeevat, 2008). is is why I take this analysis as a point of
departure for my own analysis and present it briefl y in what follows.
Onea and Volodina ( 2009 ) provide a unifi ed analysis of the various readings
of nämlich that have been attested in previous research. As they point out,
nämlich is usually analysed as having two distinct readings. On its fi rst
reading, illustrated in (1), nämlich has a specifying meaning and is used in
elliptical or parenthetical sentences.
(1) Peter hat ein Buch gekauft, nämlich “Krieg und Frieden”.
“Peter bought a book, namely ‘War and Peace’”.
In the second case, nämlich may have a causal/explanatory reading and is
syntactically an adverb positioned in the middle fi eld (MF) of the German
sentence ((2)) or in the so-called post-initial position as in (3).
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 35
e unitary semantics for nämlich that Onea and Volodina ( 2009 ) suggest is
that: (i) nämlich marks the existence of an implicit question about the preced-
ing utterance; (ii) the nämlich -sentence provides a (short) answer to this
implicit question; (iii) crucially, constituents are good short answers to con-
stituent-questions ( Who? ), while whole clauses are only good short answers to
“sentence”-questions like Why p ?
is is how the analysis works. In the case of the specifying nämlich ,
the question that the nämlich -host answers is a constituent question
whose answer specifi es a less specifi c constituent in the preceding sentence,
cf. (4).
(2) Peter hat ein Buch gekauft, er wollte nämlich was lesen.
“Peter bought a book, he wanted to read something”.
(3) Der Dieb konnte nicht fl iehen. Der Inspektor nämlich war schneller.
“ e thief couldn’t get away. e inspector was faster”.
(4) Peter hat ein Buch gekauft, nämlich “Krieg und Frieden”.
“Peter bought a book, namely ‘War and Peace’”.
Question : Which book did Peter buy?
Answer : “Krieg und Frieden”
In the case of the causal nämlich , the nämlich -host can only be seen as repre-
senting a short answer to an implicit question of the type Why ? or How come ?,
cf (5).
(5) Peter hat ein Buch gekauft, er wollte nämlich was lesen.
“Peter bought a book, he wanted to read something”.
Question : Why did Peter buy a book?
Answer : He wanted to read something
Onea and Volodina argue that the host of an MF- nämlich can answer no other
types of questions other than Why -questions, which are assumed to be the
most salient type of question that a sentence in a discourse may answer. is
amounts to assuming that nämlich cannot participate in other types of dis-
course relations other than causal explanation, an assumption that will be
falsifi ed in the next section.
In the post-initial position, nämlich is also traditionally assumed to have a
causal meaning. In this position, it additionally exhibits a particular discourse
eff ect that has been described in Breindl (2008) in terms of topic shift . is
eff ect is responsible for the contrast between (6a) and (6b) in terms of the
acceptability of nämlich in the post-initial position.
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
36 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
Onea and Volodina try to capture the topic shift eff ect of nämlich in post-
initial position in their analysis with a tentative solution along the following
lines: in reconstructing the implicit question to which the nämlich host gives
an answer, the topic to which nämlich is attached is being lifted out of the
utterance and represents a kind of hanging topic ( à la Frey, 2004 ). e implicit
question that the nämlich -sentence answers is then for a sentence like (7) Why,
with respect to yesterday, p ?, where p represents the preceding utterance.
(6) a. Die Besprechung
Topik1 wird wegen des Direktors vertagt. Der Direktor Topik2 nämlich
kann (nämlich) nicht kommen.
“ e meeting will be postponed because of the director. e director can-
not come”.
b. Der Direktor
Topik1
hat die Besprechung vertagt. Der Direktor
Topik1 ?? nämlich kann
(nämlich) nicht kommen.
“ e director has postponed the meeting. He ?? cannot come”.
In Onea and Volodina ( 2011 ), the analysis of nämlich is refi ned in a way that
I will address in the next section. e basic assumptions however remain the
same as in Onea and Volodina ( 2009 ).
While I believe the question-based solution proposed in Onea and Volodina
goes in the right direction, there are several aspects of the analysis that need
further exploration and adjustment, among which the most important is the
issue concerning the exact nature of the question that the nämlich -host can be
assumed to answer. Related to this issue is the question of the discourse rela-
tions established between the nämlich -host and its predecessor in discourse, as
well as the role of information structure in reconstructing the appropriate
implicit question.
Further, the rather poorly understood post-initial use of nämlich and its
alleged topic-shifting eff ect needs to be further explored, the more so since
topic shift is not a necessary condition for the use of nämlich in this position
as there are cases where no topic shift is involved.
In what follows, I explore the kinds of implicit discourse questions to which
the nämlich -host may be seen to provide an answer, as well as which discourse
relations are compatible with the use of nämlich . I also take a closer look at the
procedure by which an implicit discourse question (also called quaestio ) is
reconstructed. It is in the very nature of a quaestio that it is reconstructed
from a sentence in a discourse that has a particular information structure (cf.
e.g. Umbach, 2005 ). I.e., a quaestio should always refl ect the information
structure of the respective sentence in its surrounding discourse. In their
attempt to explain the topic shift eff ect related to post-initial nämlich, Onea
(7) Gestern nämlich hat es geregnet.
“Yesterday it rained”.
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 37
and Volodina ( 2009 ) follow this principle in reconstructing the quaestio for
this use of nämlich . ere, the quaestio for the post-initial nämlich is recon-
structed on the basis of the assumption that the constituent before the particle
is topical. However, this principle has been ignored in the other relevant use
of nämlich , the middle-fi eld- nämlich, the quaestio for which is assumed to be
a general Why -question. As I attempt to show, however, handling the quaestio
issue with more care may open up a way for refi ning the semantic analysis of
nämlich and improve the understanding of its discourse eff ects, especially
when it comes to the post-initial nämlich .
3. Nämlich, Implicit Questions and Discourse Relations
In this section I take a closer look at the type of discourse question that a
nämlich -sentence may answer in its so-called “causal” use. Onea and Volodina
take the stance that a Why - or How come -question is the only plausible one for
nämlich . A more thorough look at the data reveals however that while this
often is the case, there are equally frequent, clearly non-causal uses of nämlich
in the MF and post-initial position that cannot be treated within the account
of Onea and Volodina. A further observation that can be made on the basis of
corpus data is that post-initial nämlich may also have a specifi cational reading,
and that nämlich is compatible with certain kinds of causal relations and
incompatible with others. ese fi ndings cast some doubts both on the
adequacy of the traditional view on adverbial nämlich as having a causal read-
ing and on the analysis presented in Onea and Volodina. A further look at the
data reveals that there are also certain restrictions with respect to the quaestio
of a nämlich -sentence. us, a nämlich -sentence cannot answer a What hap-
pened then? or a yes/no question. Restrictions of this kind are important to
explore since they could allow deeper insights into the meaning and discourse
function of nämlich .
Since discourse questions as a means for investigating discourse structure
correlate with the discourse relations that obtain between sentences in dis-
course (e.g. Why (Explanation) What happened then ? (Narration)), a closely
related issue that will be pursued here is the question of which discourse rela-
tions nämlich may mark.
I consider fi rst the MF- nämlich . is use of nämlich is capable of indicating
answers to not only Why -questions, but also to questions that ask for specify-
ing various aspects of the state of aff airs or event described by the sentence, as
well as more abstract aspects of the utterance such as its speech act. While this
specifying function of nämlich is already a recognized fact for the elliptic uses
of nämlich , it seems not to have been observed for the so-called “causal” uses
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
38 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
1 From the Oslo Multilingual Corpus (OMC), a parallel corpus of German, English and
Norwegian texts and their translations developed at the University Oslo. e English glosses are
taken from the original translations.
2 is example is from the IDS Corpus. e English glosses are mine.
of nämlich in the MF (and neither for the post-initial nämlich ). Consider (8),
a corpus example
1 where MF- nämlich is used in the same specifi cational func-
tion as the elliptic nämlich in (4).
(8) und da hat die Gruyten was ganz Großartiges geschrieben: sie hat nämlich eine fl am-
mende Verteidigung des Grafen F… geschrieben
“and the Gruyten girl wrote something really splendid: what she wrote was an ardent
defence of Count F”. (OMC)
Here, the nämlich -clause provides a specifi cation of the non-specifi c phrase
“etwas ganz Großartiges”, thus answering the same question as the preceding
clause “What did Gruyten do/write”. While the example in (8) is less interest-
ing since it represents so to speak an expanded version of the elliptic nämlich ,
the example in (9) clearly makes the point, since here the nämlich -sentence
provides a specifi cation of the VP entsprechende Vorkehrungen treff en in the
preceding sentence.
2
(9) Vielleicht spukt es im Schloß Mespelbrunn nur deshalb nicht, weil die Vorfahren
der Ingelheims entsprechende Vorkehrungen getroff en haben: In der Decke des
Himmelbetts, das im Fürstenzimmer steht und aus dem 17. Jahrhundert stammt, sind
nämlich Fratzen geschnitzt. (IDS)
“Maybe there are no ghosts in the Mespelbrunn castle only because the ancestors
of the Ingelheims took the respective measures: in the ceiling of the 17th century
canopy bed in the room of the prince they carved grotesque faces”.
In (9), nämlich may also be associated with a causal reading, but this seems to
be due to the fact that the VP for which the nämlich -clause provides a further
specifi cation is embedded in a causal sentence. But also in (10) which lacks a
causal reading, the nämlich -sentence provides a specifi cation of the “interest-
ing consequences” that the enforcement of the bill could have.
(10) Man werde die Vorlage notfalls mit Hilfe einer Sonderverordnung durchsetzen, was
interessante Konsequenzen haben könnte. Prinz Charles hatte nämlich angekündigt,
er werde auswandern und für den Rest des Lebens Ski fahren, falls man seine geliebte
Fuchsjagd verbiete. (IDS)
“ ey would enforce the submission if necessary with the help of a special regulation,
which could have interesting consequences. Prince Charles had announced that
he would emigrate and go skiing for the rest of his life, if they forbid his beloved
fox hunting”.
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 39
3 Note that eines here is exhaustive, whereas was in (8) is not, which is related to the nature of
the specifi cation provided by nämlich that was noted in Onea and Volodina ( 2011 ) in relation
to their example (21). I will however not dwell upon the exact nature of the specifi cation pro-
vided by nämlich , leaving this issue for further research.
ese examples show two things: (i) that MF- nämlich not always has a causal
interpretation and (ii) that the specifi cational reading of nämlich is not reserved
for its elliptic uses. ey further show that nämlich may introduce a specifi ca-
tion of various less specifi c constituents of the preceding sentence, including
NPs and VPs. What is common to these examples is that the nämlich -sentence
introduces an entity or event that represents a more specifi c version of an
entity or event that is introduced in the preceding sentence and that is rather
general and non-specifi c. is non-specifi c item can often be characterized in
terms of a cataphora, especially in the elliptic cases. Consider e.g. (8), where
the cataphoric indefi nite pronoun was is used in the preceding sentence, as
well as (11), where the anaphoric indefi nite pronoun eines is used.
3
(11) Von Afrikanern, die diesen Trick durchschaut haben, kann man unterdessen hören,
daß es nur eines gebe, was schlimmer sei, als von Multis ausgebeutet zu werden,
nämlich : nicht von ihnen ausgebeutet zu werden.
“Africans who have seen through this trick have said that there is only one thing worse
than being exploited by the multinationals, and that is not being exploited by them”.
(OMC)
But we also fi nd cases where the item that the nämlich -sentence specifi es is not
explicitly non-specifi c or cataphoric, as in (12), which indicates a certain free-
dom of the speaker in deciding on which elements of the utterance need fur-
ther specifi cation. I will return to this issue towards the end of this section.
(12) Auf ihren “Stammbaum”, nämlich eine Esche, können jetzt die Abschlussklassen der
Graf-von-Oberndorff -Schule und Pestalozzi-Schule verweisen.
“ e graduating classes of the schools Graf-von-Oberndorff and Pestalozzi can now
refer to their ‘family tree’ that is, an ash”.
Cases like (8), (9) and (10), as well as (11) and (12) are reminiscent of exam-
ples of a special kind of Elaboration called Particularization that has been
extensively studied by Danlos (e.g. 1999, 2001, 2003) and that involves event
coreference, as exemplifi ed in (13).
(13) Fred damaged a garment. He stained a shirt.
Danlos ( 2003 ) argues that examples like that involve event coreference and
distinguishes between two types of discourse that involve event coreference:
particularizing and generalizing discourses. Event coreference refers to the case
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
40 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
where both sentences refer to the same event and where each element in
the second sentence S2 stands in a lexico-semantic relation (hyponymy,
hyperonymy, anaphor or identity)
4 with a corresponding element in the fi rst
sentence.
It seems that nämlich in (8)-(12) marks precisely this relation of Particu-
larization. Consider also (14) which I borrow from Danlos ( 2003 ) and trans-
late into German using nämlich . Here also, the nämlich -sentence is clearly not
in a causal relation to the preceding one. Rather, the question that the nämlich-
clause answers concerns the way in which the action described by the preced-
ing clause was performed.
4 Danlos ( 1999 ) distinguishes further between event coreference on lexical grounds, as in
(13), and event coreference based on extralinguistic knowledge.
5 From Hovy ( 1990 ).
6 From Lascarides and Asher ( 1991 ).
(14) Fred hat ein Kleidungsstück beschädigt. Er hat nämlich ein Hemd bekleckert.
“Fred damaged a garment. He stained a shirt”.
# Why did Fred do that?
What did Fred more specifi cally do?
Danlos argues further that Particularization and Generalization are more
special cases of the Elaboration relation, which have to be distinguished
from other cases of Elaboration that do not involve event coreference. Such
instances of Elaboration are cases where the second sentence introduces
further attributes of some object introduced in the fi rst sentence, as in (15),
5
or where the event in S2 contributes to the culmination of the event in S1 as
in (16).
6
(15) Knox is en route and [this ship] is a tanker.
(16) e council built the bridge. e architect drew up the plans.
Translating these examples into German, it can be observed that nämlich is
either not appropriate or changes the original reading of the sentence. Consider
(17a) where the second sentence adds to the known attributes of the ship
Knox the property of it being a tanker. Danlos treats cases like that in terms of
a relation called List. us the second sentence in (17a) adds a new item to the
list of properties of Knox.
(17) a. Knox ist unterwegs. Dieses Schiff ist ein Tanker.
b. Knox ist unterwegs. Dieses Schiff ist nämlich ein Tanker.
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 41
e situation is diff erent in (17b) where obviously due to the presence of
nämlich the List interpretation is replaced by a causal one. e discourse is
awkward since there is no plausible causal relation between being en route
and being a tanker. What this example shows is, however, that we can’t use
nämlich to add further properties of an entity on a list, i.e. that this kind
of Elaboration relation is not among the ones compatible with nämlich . But it
also raises the question of why the interpretation of the discourse with nämlich
is causal and not specifi cational? is fact suggests that the freedom of the
speaker I mentioned above to specify items of the preceding discourse is
constrained by certain principles whose nature I address towards the end of
this section.
Turning to the type of Elaboration relation found in (16), we observe that
it is not compatible with nämlich either. As (18b) suggests, nämlich changes
the interpretation, compared to (18a), towards a causal one, forcing a reinter-
pretation towards a causal Enablement relation: due to the architect, the
council was able to build the bridge.
(18) a. Der Gemeinderat hat die Brücke gebaut. Der Architekt hat die Pläne entworfen.
b. Der Gemeinderat hat die Brücke gebaut. Der Architekt hat nämlich die Pläne
entworfen.
e Elaboration relation in (18a) can also be seen as a List-type relation: it
concerns the things that are needed to build the bridge, the steps involved, one
of them being the drawing of the plans.
e discussion so far seems to suggest that discourses like Particularization
that involve event coreference are a natural environment for nämlich both in
its elliptic use and its use in MF. In such cases the nämlich sentence answers
the same implicit question as the preceding sentence but provides a more
specifi c answer to it. We also were able to detect a discourse type, involving
the List relation, which is not compatible with nämlich . Another such dis-
course is Danlos’ Generalizaton relation. Consider (19a) which represents
such a case of “generalizing restatement discourses“ involving event corefer-
ence and suggesting that staining a shirt is a more specifi c way in which one
may damage a garment in general, or that staining a shirt amounts to damag-
ing a garment. König ( 1995 ) calls this type of relation between sentences or
phrases interpretative , and Jasinskaja ( 2007 ) treats cases like this in terms of
the more general notion of Restatement. Nämlich is not appropriate in this
context, as (19b) shows. Here, the particles also (“consequently”) and somit
(“thus”) are more appropriately used, since they indicate a relation of general-
izing logical consequence and an interpretative relation between the two
sentences respectively.
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
42 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
7 e original Danlos ( 1999 ) example is “ erefore, he damaged a garment”.
Another example of Restatement is (20), where nämlich is not appropriate
either. In (20), the relation between the two sentences is not one of the second
specifying aspects of the proposition expressed by the fi rst. Rather, the second
sentence provides a more general interpretation of the state of aff airs described
by the fi rst sentence: what it means for Aljona to have lost her skis. is is also
indicated by the adverb somit (“thus”) that is appropriately used instead.
(21) a. Der einsame Reiter stieg auf das Pferd. Er ritt in den Sonnenuntergang.
“ e lone rider jumped on the horse. He rode into the sunset”.
b.
Der einsame Reiter stieg auf das Pferd. Er ritt # nämlich in den Sonnenuntergang.
“ e lone rider jumped on the horse. He rode # into the sunset”.
(20) Aljona hat ihre Schier gebrochen. Sie hat # nämlich / somit ihr Hauptverkehrsmittel
verloren.
“Aljona broke her skis. She lost her main means for transportation” (from Jasinskaja,
2007)
(19) a. Fred stained a shirt. He damaged a garment.
7
b. Fred hat ein Hemd bekleckert. Er hat # nämlich/also/somit ein Kleidungsstück
beschädigt.
“Fred stained a shirt. He /therefore damaged a garment”.
In cases of Generalizing Restatement, the two sentences address the same
implicit question, but the information provided by the second relatum is more
general than that provided by the fi rst.
So far we have identifi ed two types of discourses which are incompatible
with nämlich , List and Generalization Restatement. A further such discourse
is Narration, as in (21a), where the two sentences are related by means of a
temporal succession between the events described by them. e question that
the second sentence answers is diff erent from the fi rst and can be formulated
as What happened next?.
Looking at the restrictions for nämlich that we have observed so far, one may
conclude that nämlich is restricted to cases involving either event or object
coreference and where the nämlich -sentence provides a specifi cation of the
event or an entity described by the fi rst sentence. is would suggest an analy-
sis of nämlich in terms of Danlos’ (2003) relation of Particularization.
Unfortunately, the picture becomes more complex when one integrates the
causal uses of nämlich . Danlos ( 2001 ) considers cases of causal discourses
involving event coreference, such as the ones in (22), where a “direct causa-
tion” is involved: “the result is a physical change of state for an object Y, the
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 43
8
ese include also speech-act motivating examples such as (i):
(i) Bleibst du noch eine Minute? Ich wollte dich nämlich was fragen.
“Could you stay a minute longer? I want nämlich to ask you something”.
9
I assume that Motivation is a type of causal relation that obtains between a fact or event
and a causing attitudinal state of an intentional agent, in analogy to the reason-relation in Solstad
( 2010 ), where however a fact causes an attitudinal state. I use the term Reason for the type of
causal relation that Solstad refers to as “plain cause”, which is defi ned as a relation between a
causing fact and a caused event.
cause is an action performed by a human agent X, the action is the direct cause
of the change of state” (Danlos, 2001 : 216).
(22) a. Fred cracked the carafe. He hit it against the sink.
b. Fred hit the carafe against the sink. He cracked it.
According to Danlos, (22a) involves a relation of Explanation, whereas (22b)
a relation of Result. She argues that causal discourses which involve direct
causation diff er from other causal discourses in that they express event corefer-
ence. Examples for causal discourses that do not involve direct causation
and hence event coreference are (23a) and (23b), which express respectively
Motivation and Motivation/Narration.
(23) a. Fred broke the carafe. He was angry with Mary.
b. John cracked the carafe. Fred hit it against the sink.
Looking at corpus data, it is hard to fi nd examples of direct causation in which
nämlich occurs. e majority of the causal discourses that contain nämlich are
either Motivation
8 discourses, such as (5), where the nämlich -sentence provides
a motive for Peter for his buying a book, Justifi cation discourses such as (24a)
where the nämlich -sentence provides a justifi cation for the assertion in the
preceding sentence, or Reason-discourses such as (24b) in which the nämlich
sentence provides the cause or reason for the necessity of explaining the notion
of priority.
9
(24) a. Diese Form der inneren Notwehr ist nicht nur verständlich; sie ist unvermeidlich.
Wie eine “richtige” Reaktion auf den täglichen Massenmord aussehen sollte, weiß
nämlich niemand zu sagen.
“ is form of internal self-defence is not only understandable, it is unavoidable;
who can prescribe the ‘correct’ response to the daily slaughter?”. (OMC)
b. Vielleicht ist es nötig, zu erklären, was eine Priorität ist. Es gibt nämlich genügend
Leute, die sich dümmer stellen als sie sind, sobald ein Argument nicht in ihr
Weltbild paßt.
“Perhaps we ought to explain what a priority is. Some people pretend to be more
stupid than they really are as soon as they encounter an argument that doesn’t fi t
their view of the world”. (OMC)
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
44 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
Such Motivation, Justifi cation or Reason-discourses do not satisfy Danlos’
narrow defi nition of direct causation and hence Explanation since they do
not involve event coreference: in (5), as well as in (24a) and (24b), the
nämlich sentence and its predecessor describe two separate events. Moreover,
nämlich does not seem quite appropriate in causal Explanation contexts like
(22a). 10 Consider (25) which is a translation of (22a). Here the use of nämlich
changes the original direct causation interpretation towards one in which the
nämlich -sentence is a specifi cation of the way in which Fred broke the carafe,
i.e. a Particularization relation that also involves event coreference. e direct
causation interpretation is in contrast achieved when indem (“by”) instead of
nämlich is used (cf. (25b)).
(25) a. Fred hat die Karaff e zerbrochen. Er hat sie nämlich gegen die Spüle gestoßen.
“Fred broke the carafe. He hit it against the sink”.
b. Fred hat die Karaff e zerbrochen, indem er sie gegen die Spüle gestoßen hat.
“Fred broke the carafe by hitting it against the sink”.
On the other hand, Result-discourses are completely inappropriate with näm-
lich , as (26) shows, which is not unexpected, since Result discourses are the
causal pendant of Generalization discourses, as argued in Danlos ( 2001 ).
10 I owe this remark to an anonymous reviewer of an earlier draft of this article.
(26) Fred hat die Karaff e gegen die Spüle gestoßen. Er hat sie #nämlich zerbrochen.
“Fred hit the carafe against the sink. He broke it”.
Summing up, a more thorough examination of corpus data suggests that näm-
lich is compatible with (i) either non-causal Particularization discourses that
involve event coreference or (ii) causal discourses that do not involve event
coreference. It also suggests that not only elliptic but also MF- nämlich may
participate in non-causal Particularization discourses. ese fi ndings give a
more refi ned picture of the empirical data compared to the one presented in
Onea and Volodina ( 2009 ) and Onea and Volodina ( 2011 ), where the analysis
accounts for MF- nämlich as indicating a short answer to Why -questions only
and rules out “discourse subordinating but at the same time topic changing
questions such as elaborative questions”. is is a bit unfortunate since, as
I have attempted to show, not all cases of Elaboration involve the type of ques-
tions envisaged by Onea and Volodina ( 2011 ), e.g. those in (17) and (18), and
since the cases of Elaboration compatible with nämlich are exactly the cases in
which the question is not a “topic changing” one but one that is identical with
the implicit question of the preceding sentence.
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 45
is means that examples like (10) would be treated by Onea and Volodina
as causal, which contradicts the intuitions.
In spite of these empirical issues, Onea and Volodina arrive at a unifi ed
specifi cational meaning of nämlich, which is compatible with the general
proposal that this article makes. Onea and Volodina capture the meaning of
nämlich in terms of it indicating a short answer (and thus specifi cation) of an
implicit question about the preceding sentence, assuming that the causal read-
ings arise in the process of pragmatic interpretation. e lexical entry assumed
in Onea and Volodina ( 2011 ) improves the one in Onea and Volodina ( 2009 )
by replacing the problematic assumption of a salience hierarchy of questions:
Why is a Why -question more salient than a What happened next -question?
Usually, Narration is taken to be the default discourse relation by a QUD-
(question under discussion) mechanism, which ensures that the nämlich -
sentence provides a complete answer to a salient implicit question about the
preceding utterance. is account presupposes that the question about the
preceding utterance that the nämlich -sentence answers is already on the QUD-
stack, that is, that the speaker has a complete strategy consisting of questions
and subquestions to be addressed in the course of the conversation. us the
felicitous use of nämlich in (12) would be accounted for by the fact that the
constituent introduced by nämlich represents a short answer to the question
“What kind of tree is that?” that the speaker has on his QUD-stack. After
providing the necessary information, namely the complete answer to this
question, the question is removed from the stack. is is argued to capture the
intuition that the speaker “is specifying what he meant by his own previous
utterance” and that this is in “his sole competence”.
While this seems to be a plausible argument, the data I discussed suggests
that at least in some cases, a diff erent question-based analysis may be suggest-
ing itself, especially in the cases involving coreference, where intuitively the
implicit discourse question that the nämlich -sentence answers is the same as
the one that the preceding sentence answers, rather than “about the preceding
utterance”: consider again (12) where the question that both the sentence and
the elliptic nämlich -construction answer is “What can the graduating classes
refer to?”, the answer provided by the elliptic nämlich being more specifi c than
the one provided by the matrix sentence.
I would therefore like to take a diff erent approach and suggest that nämlich
signals a specifi cational relation between a sentence constituent or a sentence
and the preceding sentence in discourse. I suggest further that the causal cases
can be reconciled with the specifi cational ones by viewing causes as specifi cations
of their eff ects. Asserting only the result of some causal process is less specifi c
than asserting also the cause for this result. On the other hand, the mere
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
46 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
temporal succession of two events or the mere listing of states or events cannot
be seen as a case of specifi cation of the event described by the preceding sen-
tence. In Narration or List discourses, the second sentence does not specify the
fi rst in any respect. In the case of causal relations such as Motivation and
Reason, the cause can be seen as a specifi cation of the result. is intuition can
be captured by assuming a covert contextual cause-variable associated with each
utterance (including its speech act), which may be instantiated, by the nämlich -
sentence. at this account may be on the right track is suggested by the follow-
ing observation. While other contextual variables such as time, place, and degree
are not appropriate arguments of specifi cational nämlich , as noted by Onea
and Volodina ( 2011 ) and attested by (27a), cases like (27b) are quite natural.
(27) a. Peter küsst Maria. #Nämlich in der Küche.
b. Peter küsst Maria. Nämlich weil er sie liebt. /Weil er sie nämlich liebt.
Further, naturally occurring examples are given in (28).
(28) a. den passiert es eben, dass die Patientin das Behandlungskonzept nicht zur Gänze
durchführt, nämlich weil sie es von vornherein so nie wollte und weil sie sich am
Anfang nicht genügend ausgekannt hatte! (Google)
“it just happens that the patient does not carry out the treatment concept in its
entirety, namely because she didn’t wanted that from the outset and because she
wasn’t suffi ciently informed in the beginning!”
b. Wie Sie sagen, ein Schweigen aushalten, das, glaube ich, ist hier eher die Prob-
lematik von Frauen, nämlich weil sie glauben, sie müssten die Pausen füllen.
(Google)
“As you say, to endure silence, this I think is more a women’s issue, namely because
they think they have to fi ll the pauses”.
Taking the cause of a state or fact (or speech act) to be a further specifi able
contextual variable of each utterance fi ts nicely with the specifi cational uses of
nämlich where the nämlich -sentence often provides the material that binds an
explicit variable such as a cataphoric expression. On the other hand, the
implicit cause-variable needs not be instantiated or bound, whereas the cata-
phors do need to be specifi ed, otherwise the discourse is infelicitous since it
leaves information open that the previous discourse promises to supply. e
cases of Particularization based on a hyponymy-hyperonymy relation are
similar to the causal ones in this respect, since, although they involve rather
non-specifi c concepts, their specifi cation is not necessary for the appropriate-
ness of the discourse. But we saw that also in the elliptic cases, the preceding
context may not require specifi cation, since the concepts involved are of a
satisfactory granularity. is line of thought may also be suggesting an answer
to the question I put forward earlier in this section about what constrains the
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 47
11 anks to an anonymous reviewer for pointing this out.
freedom of the speaker to specify elements of the utterance. On the back-
ground of the data I presented, a possible ranking emerges according to which
the nämlich -sentence is preferably interpreted as providing a specifi cation of
explicitly marked non-specifi c aspects of the utterance, and a causal interpre-
tation arises only if no such elements are available. is would explain why the
discourse in (17b) gets a causal interpretation with nämlich : it does not con-
tain other specifi able elements.
Cast in a question-based framework, the above conclusions would imply
that an appropriate generalisation of the question that a nämlich -sentence
answers is a question that asks for specifying one of 3 possible non-specifi c
aspects of the utterance mentioned above. In other words, the meaning of
nämlich is more adequately captured in terms of indicating an answer to a
discourse question that asks for more detailed information about some aspect
of the state of aff airs described by the preceding sentence, rather than a
Why -question as suggested by Onea and Volodina. I call the discourse ques-
tion that a nämlich -sentence answers a specifying question and defi ne it as a
question that addresses an aspect of the utterance that is further specifi able in
the given context, including non-specifi c expressions such as cataphors and
(nonspecifi c) indefi nites or plurals, implicit causes, and hyperonyms. is list,
as well as the assumptions put forward here has to be further refi ned by more
extensive corpus studies. Since I argued that both the nämlich -sentence and
the preceding utterance answer one and the same question, the “specifying
question” I envisage here should be seen as a subquestion of this common
superquestion. e details of the analysis have yet to be worked out and cast
in a formal framework.
Concerning the exact nature of this specifi cation relation, Onea and
Volodina ( 2011 ) provide a more fi ne-grained overview of further constraints
on the specifi cation that nämlich -sentences provide. us they observe that
the specifi cations that nämlich -sentences provide have to be complete, although
there are cases in which this requirement does not fully apply, suggesting that
more work is needed to reveal the exact nature of this specifi cational relation.
Further evidence for the nature of the specifi cational relation contributed by
nämlich may be provided by examining its behaviour in relative and adverbial
subordinate clauses, as well as by examining the way nämlich combines with
other discourse particles that signal discourse relations of their own.
11 ese
issues however have to be postponed for future work.
I now turn to the case of post-initial nämlich and try to fi nd out whether it
obeys the same restrictions as the MF- nämlich with respect to the discourse
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
48 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
12 Another possible structuring that licenses postinitial nämlich is Fred hat ein Kleidungsstück
beschädigt. Ein Hemd nämlich hat er bekleckert .
relation that the nämlich -sentence is compatible with. At fi rst glance, this
nämlich does not seem appropriate in cases of Particularization discourses
such as (29a), while it is appropriately used in cases of Reason-discourses such
as (29b).
(29) a. Fred hat ein Kleidungsstück beschädigt. Er # nämlich hat ein Hemd bekleckert.
“Fred has damaged a garment. He has stained a shirt”.
b. Der Dieb konnte nicht fl iehen. Der Inspektor nämlich war schneller.
“ e thief couldn’t escape. e Inspector was faster”.
(30) Ein Kleidungsstück wurde beschädigt. Ein Hemd nämlich wurde bekleckert.
“A garment was damaged. A shirt was stained”.
However, a more thorough look reveals that the reason for the ill-formedness
of (29a) is not the obtaining relation of Particularization but rather the way
the nämlich -sentence is informationally structured. With a diff erent structur-
ing, nämlich is perfectly acceptable.
12
(30) shows that post-initial nämlich cannot be treated in terms of addressing
a Why-question either, contrary to what Onea and Volodina ( 2009 ) suggest. It
also suggests that restrictions of the use of nämlich in post-initial position
compared to the other uses of nämlich are not semantic but merely due to
information structure. is means that since post-initial nämlich is compatible
with the discourse relation of Particularization, an analysis of nämlich in terms
of specifi city should be capable of integrating also its post-initial use.
What remains to be seen is how an analysis of nämlich in terms of specifi city
can capture the additional topic-shift eff ect attested in cases of post-initial
nämlich . I will next show that taking into account the information structure
of the nämlich -sentence in trying to reconstruct the discourse question that it
answers opens up a possibility of accounting for the diff erence between
MF- and postinitial- nämlich in terms of the topic-shift eff ect observed in con-
nection with the latter.
4. Nämlich and the Information Structure of the Host
In the previous section it was suggested that a discourse question analysis of
nämlich in terms of indicating the answer to a Why -question is not adequate.
Such an approach is not helpful in accounting for the diff erence between
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 49
13 e bracketing that represents the focus domains in this and the following examples is only
very rough and leaves open the question of whether nämlich must be seen as part of the focus
e.g. in cases like (31).
MF- and postinitial- nämlich either. is diff erence is illustrated by (31) and
referred to as the topic-shift eff ect of post-initial nämlich. On such an
account, the question addressed by the nämlich sentences would be identical
in the two cases.
(31) a. Die Besprechung wird wegen des Direktors vertagt. Der Direktor nämlich ist krank.
“ e meeting is postponed because of the director. e director is ill”.
Why is the meeting postponed?
b.
Die Besprechung wird wegen des Direktors vertagt. Der Direktor ist nämlich krank.
“ e meeting is postponed because of the director. e director is ill”.
Why is the meeting postponed?
(32) Fred hat ein Kleidungsstück beschädigt. [Er]
T hat nämlich [ein Hemd bekleckert]
F .
“Fred damaged a garment. He has stained a shirt”.
What did Fred specifi cally do to damage a garment?
is is why Onea and Volodina ( 2009 ) suggest an analysis of postinitial-
nämlich in terms of a diff erent quaestio that refl ects the association of nämlich
with the sentence topic. Taking the information structure of the sentence into
account is a general principle when it comes to reconstructing the discourse
question that a sentence can be seen as an answer to in the particular dis-
course. It is therefore essential that this principle is applied also in the cases of
MF- nämlich . is is what I will do in what follows, since neither Onea and
Volodina ( 2009 ) or Onea and Volodina ( 2011 ) dwell on this issue. is
approach will take us eventually to an integrated analysis of nämlich in terms
of specifi cation.
In what follows, I argue that MF- nämlich and post-initial nämlich answer a
diff erent specifying question depending on the constituent nämlich associates
with and its information-structural property. Consider (32). Here, MF- nämlich
associates with the focus.
13 e discourse question that the nämlich -host
answers is a question that requires specifi cation of the event of damaging a
garment introduced by the preceding sentence. is specifi cation is given by
the VP-focus of the nämlich -host, which is also the unit to which nämlich
syntactically attaches.
e illformedness of the same sentence with post-initial nämlich can be
explained along the following lines: In “Er nämlich hat ein Hemd bekleckert”,
nämlich associates with the topic Er and thus identifi es Er as a more specifi c
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
50 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
description of some aspect of the preceding sentence. I.e., the question that
the nämlich -sentence answers is Who specifi cally damaged a garment ?. is
question however does not license a natural continuation of the discourse,
since it is already answered by the preceding sentence “Fred hat ein Klei-
dungsstück beschädigt”. In other words, an information structuring like this
is not licensed by the context.
(33) Fred hat ein Kleidungsstück beschädigt. [Er]
T # nämlich [hat ein Hemd bekleckert]
F .
“Fred damaged a garment. He stained a shirt”.
# Who specifi cally damaged a garment?
In contrast, in (34) post-initial nämlich identifi es the topic Hemd as the ele-
ment that represents a more specifi c description of some aspect of the preced-
ing sentence. e corresponding discourse question What garment was more
specifi cally damaged ? leads to a coherent continuation of the discourse.
Note that here, ein Hemd is a contrastive topic : a shirt is an element of the set
of garments. Note also that in (34), no topic shift is involved. What is involved
here can be rather called “topic specifi cation”.
A more intricate case is (35). e solution to this and similar cases that
I suggest below within the specifi cational analysis of nämlich provided in the
previous section is only tentative and needs further elaboration.
(34) Ein Kleidungsstück wurde beschädigt. [Ein Hemd]
CT nämlich [wurde bekleckert] F .
“A garment was damaged. A shirt was stained”.
What garment was more specifi cally damaged (and how)?
(35) a. Die Besprechung wird [wegen des Direktors] vertagt. [Der Direktor]
T nämlich
[ist krank]
F .
“ e meeting is postponed because of the director. e director
is ill”.
Who specifi cally is the reason for the meeting to be postponed?
b.
Die Besprechung wird wegen des Direktors vertagt. [Der Direktor]
T ist nämlich [krank] F .
“ e meeting is postponed because of the director. e director is ill”.
What specifi cally is the reason for the meeting to be postponed?
c. Der Direktor hat die Besprechung vertagt. [Der Direktor]
T # nämlich ist [krank]
F .
“ e director has postponed the meeting. e director is ill”.
# What specifi cally is the reason for the meeting to be postponed?
In (35a), the post-initial nämlich marks intuitively the director as the under-
specifi ed reason for the postponement of the meeting. is structuring is
licensed by the preceding sentence in which the causal wegen -PP already
identifi es the director as the underspecifi ed reason for the postponement of
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 51
the meeting. Here, the PP instantiates the implicit cause-variable that I argued
is related to each utterance. However, the individual that instantiates this
variable represents an incomplete, underspecifi ed reason, since reasons are
normally propositions or facts. e wegen -PP indicates therefore that the rea-
son is related to an underspecifi ed property of the director. e “missing” part
of the causing proposition is provided by the rest of the nämlich -sentence: the
property of the director being ill. us, post-initial nämlich plays a twofold
role: it indicates that the nämlich -sentence represents a specifi cation of some
parameter/aspect of the preceding sentence that is left unspecifi ed, and it
identifi es the entity denoted by the initial fi eld constituent as the entity that
has the necessary properties for providing this specifi cation. e specifi cation
that nämlich seems to provide here can be paraphrased as “ e director with
his underspecifi ed property is the reason for the postponement”.
In contrast, in (35b), where nämlich is attached to the VP, it provides the
direct specifi cation of the underspecifi ed property of the director that leads to
the postponement of the meeting. I.e. nämlich in this position identifi es the
property of the director being ill as the implicit underspecifi ed property in the
preceding sentence, and thus as the constituent that specifi es this underspeci-
fi ed property.
But why is (35c) not felicitous? Here the situation is diff erent from (35a) in
that in (35a) the director is explicitly identifi ed as the reason for the postpone-
ment of the meeting. e wegen -PP in the preceding utterance indicates that
the reason for the postponement is related to an underspecifi ed property of
the director. No such indication is involved in (35c) where the reason for the
postponement is open: it may be the director, but also his assistant, or a fi re
alarm. It seems that post-initial nämlich indicates a question about an entity
rather than a question about a property. And, in the absence of explicit
non-specifi c elements of the preceding discourse, it is a question about a rea-
son: Who/ What entity is the reason for p ?, where p is the preceding clause.
Intuitively, the nämlich -sentence in (35c) is not an appropriate answer to this
question, since the answer it would be providing could be paraphrased as “ e
director is the reason why the director has postponed the meeting”. Such an
answer is not specifi c enough, and thus contradicts the semantics of nämlich ,
namely the indication of the rendering of a more specifi c version of some
proposition. In contrast, in (35a) where the director is already identifi ed as the
reason for the postponement, nämlich provides a diff erent and, albeit mini-
mally, more informative specifi cation that I paraphrased as “ e director with
his underspecifi ed property is the reason for the postponement”. In fact,
a crucial aspect of the problem with (35a,c) seems to be the fact that the
structuring of (35a) is licensed by the fact that the director is already known
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
52 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
to be the reason for the postponement, which makes the constituent denoting
this entity an appropriate topical expression. It seems that a combination of
factors is at work here, the exact interaction of which is a question that requires
more thorough consideration than can be provided here.
A more simple case is (3), repeated below as (3’a), where nämlich marks the
inspector as the immediate reason for the fact that the thief couldn’t escape.
e rest of the sentence then completes the causing proposition. e structur-
ing is licensed by the context, since both the thief and the inspector belong to
a knowledge frame, in which inspectors (with their particular properties) are
plausible reasons for preventing thieves from escaping. e inspector counts
thus as a given entity, which is usually taken as a property of topics.
Again, a discourse like (3’b) with the thief as a topic is not acceptable, since
the entity marked as the reason for the fact that the thief was not able to escape
is not specifi c enough. On the other hand, (3’c) is felicitous, since the prop-
erty of the thief being too slow is a plausible and specifi c enough reason for
him being not able to escape.
Also cases like (36) with post-initial nämlich (from Onea and Volodina,
2009 ) can be treated along these lines. Here, the anaphoric ambiguity involved
is resolved in favour of the object “the son” rather than the subject (and topic)
“the father”, which suggests that at least in cases like that, nämlich clearly
requires a topic shift.
(3’) a. Der Dieb konnte nicht fl iehen. Der Inspektor nämlich war schneller.
“ e thief couldn’t get away. e inspector was faster”.
b. Der Dieb konnte nicht fl iehen. Der Dieb/Er # nämlich war zu langsam.
“ e thief couldn’t get away. e thief/He was too slow”.
c. Der Dieb konnte nicht fl iehen. Der Dieb/Er war nämlich zu langsam.
“ e thief couldn’t get away. e thief/He was too slow”.
(36) a. Der Vater geht zu seinem Sohn. Er (Der Sohn) nämlich ist krank.
‘ e father goes to his son. He ( e son) is ill.’
b. Der Vater geht zu seinem Sohn. Er (#der Vater) nämlich ist krank.
‘ e father goes to his son. He ( e father) is ill.’
An attempt to explain this case draws an analogy between (36) and (3’).
In (3’), der Dieb is not a suffi ciently specifi c reason for the thief not being able
to escape, but his being too slow is. Similarly, “the father” in (36b) is not a
suffi ciently specifi c or informative reason for the father going to the son,
whereas “the son” is more informative and can be seen as a more specifi c rea-
son for the father going to his son.
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 53
14 e example is due to Cornelia Ebert.
is means that post-initial nämlich is licensed in contexts in which the
subject of the nämlich -sentence can be seen as a plausible, underspecifi ed rea-
son for the eventuality described in the preceding sentence, or if the subject is
already identifi ed as the reason, as in (35a). is is however a tentative hypoth-
esis that needs to be verifi ed in further research.
e observations made so far suggest that nämlich is neither a focussing nor
a topic particle (contra Breindl 2008 ) but that it merely is capable of attaching
to or associating with a constituent that is identifi ed as a more specifi c descrip-
tion of some aspect of the preceding sentence in discourse. Its function can be
thus described in terms of marking or identifying the constituent that repre-
sents a more specifi c version of an element of the preceding sentence. is
suggests in turn that topic shift is not a necessary eff ect related to the post-
initial position of nämlich but that whether nämlich may be used in this posi-
tion or not depends on the properties of the preceding context and more
closely on the aspects of the sentence that are further specifi able. If this
happens to be a constituent that may occupy the initial position, then nämlich
can be appropriately attached to it and thus used in the post-initial position.
In what follows I show how the specifi cation analysis of nämlich may deal with
cases of post-initial nämlich that do not involve topic shift.
I already considered one such example earlier, (34), which I argued does not
involve topic shift but rather what can be called a topic specifi cation. Another
case with post-initial nämlich that does not involve topic shift is (37),
14 where
Er in the nämlich -sentence is not a topic in the sense of an aboutness topic,
but rather a contrastive topic (CT) involving exhaustivity, since the implicit
topic alternatives are denied. e sentence is most appropriately realised with
a contrastive accent on Er , rather than with no accent. is is also suggested
by the salience of a set of persons triggered by the expression “the only one” in
the preceding sentence that licenses a contrastive reading on which “Er” is
contrasted with the rest of the set of salient persons.
(37) Peter ist der einzige, der nett ist. [Er]
CT nämlich [bringt immer Schokolade mit].
“Peter is the only one who is nice. He always brings chocolate”.
Intuitively, (37) provides a defi nition of what it means to be nice according to
the speaker. It says that only Peter is nice because he is the only one who brings
chocolate. is reading is not available in the case of MF- nämlich : (38)
suggests a Motivation interpretation in terms of the reason for calling Peter
the only one who is nice.
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
54 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
15 anks to an anonymous reviewer for this remark.
(38) Peter ist der einzige, der nett ist. Er bringt nämlich immer Schokolade mit.
“Peter is the only one who is nice. He always brings chocolate”.
It seems that the “defi nitional” reading in (37) is achieved by identifying Peter
and his unique, not yet specifi ed property, (where the uniqueness is suggested
by the contrastive accent and the property is specifi ed by the remaining sen-
tence), as the reason for calling Peter “the only one who is nice”. e sugges-
tion that Peter’s property or habit of bringing chocolate is unique is the crucial
ingredient that leads to the defi nitional reading: the only one who is nice is
identifi ed with the only one who brings chocolate, and this is Peter. e answer
that the nämlich -sentence in (37) provides is thus “Peter with his unique prop-
erty is the reason for calling him P, and this property is Q”. e defi nitional
reading involved in (37) is also licensed when the uniqueness or exhaustivity
is indicated by other means, as in (39).
15
(39) Peter ist der einzige der nett ist. Er ist nämlich der einzige, der immer Schokolade
mitbringt.
“Peter is the only one who is nice. He is nämlich the only one who always brings
chocolate”.
is analysis of (37) is similar to the one I suggested for (35a), where the
wegen -PP licenses an interpretation on which the entity that the topical expres-
sion denotes is identifi ed as the reason for the resulting event, due to an under-
specifi ed causing property. In (37), the same eff ect is achieved by exhaustivity
due to contrastive accent: exhaustivity licenses an interpretation on which the
entity that the topical expression denotes is identifi ed as the reason for the
speech act, due to an underspecifi ed, unique motivating property of the topi-
cal subject.
In contrast, on an analysis in which Er is an aboutness topic, no unique
causing property of the topical entity would be suggested and the implicit
question would be reconstructed as Who specifi cally is the one who is nice ?, which
would be inappropriate in this context since this question is already answered
by the fi rst sentence. I.e., Peter as a sentence topic is not further specifi able in
this context: without the contrastive context the post-initial nämlich is not
appropriate since it signals that the sentence answers a question that leads to a
non-coherent continuation of the discourse. Consider also (40) where com-
pared to (37) the contrast with the rest of the salient set of persons is lacking
and where Er cannot be interpreted as contrastive topic. Here, post-initial
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 55
16 Also due to C. Ebert.
nämlich is not appropriate since also here no unique causing property of the
topical entity is suggested and the discourse question that corresponds to the
nämlich -sentence with such a structuring is not licensed by the context.
(40) Peter ist nett. Er ?? nämlich bringt immer Schokolade mit.
“Peter is nice. He always brings chocolate”.
# Who specifi cally is nice ?
Similarly, in (41), which lacks the contrast to the set of salient persons,
MF- nämlich is interpreted along the same lines as an indication that the sen-
tence answers the discourse question What is the specifi c reason for calling Peter
nice?. In both (38) and (41), nämlich is attached to the VP and this identifi es
or specifi es the act of “bringing chocolate” as doing something nice.
(41) Peter ist nett. Er bringt nämlich immer Schokolade mit.
“Peter is nice. He always brings chocolate”.
(42) Ich gehe heute gar nicht an die Uni. [Peter] nämlich möchte ich auf keinen Fall treff en.
What is the specifi c reason for not wanting to go to the university?
Another case with post-initial nämlich that does not involve topic shift is
(42). 16 On my analysis, post-initial nämlich is appropriately used here because
it marks that the constituent it associates with, Peter (which is not an about-
ness topic), is the specifi c reason for the speaker not wanting to go to the
university.
Summing up, I argued that whether nämlich is appropriately used in
post-initial position depends on whether the fi rst constituent that nämlich
associates with is an entity that represents a more specifi c aspect of the state of
aff airs described by the preceding sentence or not. Depending on the particu-
lar context, this may be accompanied by a topic shift, but topic shift is not a
requirement for using nämlich in this position, nor does it seem to be an eff ect
of the specifi c position. is explains why post-initial nämlich is not always
accompanied by topic shift, as well as why the fi rst constituent may be not
only aboutness topic (which is usually associated with the process of topic
shifting) but also a contrastive topic or not a topic at all.
e contribution of nämlich can be specifi ed in terms of explicitly marking
a relation of specifi cation between its host and the sentence that precedes it.
It indicates that its host provides a more specifi c description of some (type
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
56 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
compatible) entity introduced in the fi rst sentence or some implicit “specifi -
able” aspect of it (or the speech act it realises), such as the reason for the state
of aff airs described by it. e position of the particle identifi es the entity that
represents this more detailed description: an entity (postinitial), or property
(MF). e details of the analysis presented here are yet to be spelled out in
more precise terms.
A problem for this analysis that remains to be addressed is related to the
object type of the entity denoted by the constituent with which post-initial
nämlich is associated. As was pointed out by an anonymous reviewer, these
may be constituents that denote not only individuals but also times. A corpus
search revealed that this position may be occupied by spatial and temporal
pronouns such as Dort (“there”) and dann (“then”), demonstrative pronouns
such as das (“this”), time specifi cations such as Minuten später (“minutes later”)
and conditional conjunctions such as dann (“then”). While this data poses a
problem for an analysis in terms of reasons, there is some indication that such
cases may be analysed in terms of noncausal Particularization. is issue has to
be, however, postponed for further work.
5. Summary and Conclusions
I have argued that the meaning of the German discourse particle nämlich is
best captured in terms of a relation of specifi cation between two sentences.
More specifi cally, I have argued that its second argument is a specifi cation of
some (type compatible) entity in its fi rst argument or the reason of the state of
aff airs described by the fi rst sentence, and that the position of the particle
identifi es the element of the preceding sentence that is specifi ed by the näm-
lich -sentence. I have also argued that the discourse relations that nämlich is
compatible with are relations that not simply involve event coreference but
that involve some specifi cation of some underspecifi ed aspects of the utter-
ance, such as certain types of elaboration and causal relations. e discourse
eff ects of post-initial nämlich seem to be related to the ability of nämlich
to associate with diff erent constituents belonging to diff erent information-
structural units and thus to address diff erent discourse questions. e rela-
tion between the meaning of nämlich and discourse eff ects like topic shift/
continuity depends on the aspects of the preceding utterance that are further
specifi able in the particular context. e fact that nämlich , in contrast to other
discourse particles such as deshalb (“hence”) and somit (“thus”) is capable of
occupying the post-initial position is related to its ability to specify aspects of
the preceding utterance that can be realized in the topic position.
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58 57
e account presented here is rather informal and needs to be made consid-
erably more precise, which is an enterprise that will be pursued in further
research. It also remains to be seen how the presented analysis of nämlich
relates to other discourse particles capable of occupying the post-initial posi-
tion, such as aber , hingegen and schließlich, especially with respect to the ques-
tion of how their distribution is related to their meaning. A further important
question that remains to be answered is whether the topic shift eff ect that has
been observed with other particles in the post-initial position can also be seen
as a side eff ect, or whether this eff ect is a property of the syntactic position and
the case of nämlich should be seen as an exception. ese are however ques-
tions that need to be postponed for future work.
References
Beaver , David and Brady Clark . 2008 . Sense and Sensitivity: How Focus Determines Meaning .
Malden, MA : Blackwell.
Breindl , Eva . 2008 . ‘Die Brigitte nun kann der Hans nicht ausstehen’. Gebundene Topiks im
Deutschen . In E. Breindl and M. urmair (eds.), Erkenntnisse vom Rande. Zur Interaktion
von Prosodie, Informationsstruktur, Syntax und Bedeutung . Zugleich Festschrift für Hans
Altmann zum 65. Geburtstag , 27 - 49 . Berlin : Erich Schmidt Verlag.
Danlos , Laurence . 1999 . Event coreference. Between two sentences . Proceedings of the
International Workshop of Computational Semantics , 15 - 33 . University of Tilburg .
Danlos , Laurence . 2001 . Event coreference in causal discourses . In P. Bouillon and F. Busa (eds.),
e Language of Word Meaning , 216 - 241 . Cambridge : Cambridge University Press .
Danlos , Laurence . 2003 . Event coreference between two sentences . In H. Bunt , R. Muskens and
E. ijsse (eds.), Computing Meaning Vol. 2 , 271 - 288 . Dordrecht : Kluwer .
Frey , Werner . 2004 . Notes on the syntax and the pragmatics of German Left Dislocation .
In H. Lohnstein and S. Trissler (eds.), e Syntax and Semantics of the Left Periphery ,
203 - 233 . Berlin : Mouton de Gruyter .
Granito , Marius . 1984 . ‘Nämlich’, ‘und zwar’: étude syntactique et sémantique . Cahiers d’Études
Germaniques 8 : 165 - 209 .
Hartmann , Dietrich . 1977 . Aussagesätze, Behauptungshandlungen und die kommunikativen
Funktionen der Satzpartikeln ja , nämlich und einfach . In H. Weydt (ed.), Aspekte der
Modalpartikeln. Studien zur deutschen Abtönung , 101 - 114 . Tübingen : Niemeyer .
Hovy , Edward . 1990 . Unresolved issues in paragraph planning . In R. Dale , C. Mellish and
M. Zock (eds.), Current Research in Natural Language Generation , 17 - 45 . New York :
Academic Press .
Jasinskaja , Katja . 2007 . Pragmatics and Prosody of Implicit Discourse Relations . PhD dissertation ,
University of Tübingen.
Jasinskaja , Katja and Henk Zeevat . 2008 . Explaining additive, adversative and contrast marking
in Russian and English . Revue de Sémantique et Pragmatique, 24: 65-91 .
König , Ekkehard . 1995 . e meaning of converb constructions . In M. Haspelmath and
E. König (eds.), Converbs in Cross-Linguistic Perspective , 57 - 96 . Berlin : Walter de Gruyter .
Lascarides , Alex and Nicholas Asher . 1991 . Discourse relations and defeasible knowledge .
Proceedings of the 29th Annual Meeting of the Association of Computational Linguistics ,
36 - 47 . University of California , Berkeley.
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
58 E. Karagjosova / International Review of Pragmatics 3 (2011) 33–58
Onea , Edgar and Anna Volodina . 2009 . Der Schein trügt nämlich . Linguistische Berichte
219 : 291 - 321 .
Onea , Edgar and Anna Volodina . 2011 . Between Specifi cation and Explanation: About a
German Discourse Particle . International Review of Pragmatics 3: 3-32 .
Roberts , Craige . 1996 . Information structure in discourse. Towards an integrated formal theory
of pragmatics . In J.-H. Yoon and A. Kathol (eds.), Ohio State University Working Papers in
Linguistics , 91 - 136 . Department of Linguistics, Ohio State University .
Solstad , Torgrim . 2010 . Some new observations on ‘because (of)’ . In M. Aloni and K. Schulz
(eds.), Amsterdam Colloquium 2009. Lecture Notes in Computer Science , 60 - 77 . Berlin :
Springer .
Umbach , Carla . 2005 . Contrast and information structure: A focus-based analysis of ‘but’ .
Linguistics 43 : 207 - 232 .
Vinckel , Hélène . 2008 . ‘Nämlich’ vs. ‘und zwar’? Konkurrenz oder Komplementarität?
In D. Baudot and M. Kauff er (eds.), Wort und Text. Lexikologische und textsyntaktische
Studien im Deutschen und Französischen. Festschrift für René Métrich zum 60. Geburtstag ,
341 - 350 . Tübingen : Stauff enburg Verlag .
Downloaded from Brill.com03/26/2020 08:10:47PM
via FU Berlin
