ArticlePDF Available

Abstract and Figures

In this paper we argue that the notions generally grouped together under the heading of evidentiality actually belong to four different evidential subcategories, which are different from one another in terms of their semantic scope. The hierarchical, scopal architecture of Functional Discourse Grammar is used to define these four categories. After giving our arguments for this new classification, we test a number of predictions that follow from it concerning the coexistence of evidential subcategories within a language and the co-occurrence of evidential markers in a single clause. We investigate our predictions in a sample of 64 native languages of Brazil. The data from these languages show that the presence of one or more of the four evidential subcategories can be systematically described in terms of an implicational hierarchy.
Content may be subject to copyright.
Linguistics ; 53(3): 479–524
Kees Hengeveld* and Marize Mattos Dall’Aglio Hattnher
Four types of evidentiality in the native
languages of Brazil
Abstract: In this paper we argue that the notions generally grouped together
under the heading of evidentiality actually belong to four dierent evidential
subcategories, which are dierent from one another in terms of their semantic
scope. The hierarchical, scopal architecture of Functional Discourse Grammar
isused to dene these four categories. Aer giving our arguments for this new
classication, we test a number of predictions that follow from it concerning the
coexistence of evidential subcategories within a language and the co-occurrence
of evidential markers in a single clause. We investigate our predictions in a sam-
ple of 64 native languages of Brazil. The data from these languages show that
thepresence of one or more of the four evidential subcategories can be systemat-
ically described in terms of an implicational hierarchy.
Keywords: evidentiality, event perception, deduction, inference, reportativity,
Functional Discourse Grammar, native languages of Brazil
DOI ./ling--
Introduction
In this paper we argue that the notions generally grouped together under the
heading of evidentiality actually belong to four dierent evidential subcatego-
ries, diering from one another in terms of their scope. The hierarchical, layered
architecture of Functional Discourse Grammar (Hengeveld and Mackenzie 2008)
will be used to dene these subcategories as applying to dierent hierarchical
layers within the grammar and is explained in Section 2. The new and fourfold
classication of evidential categories resulting from this approach is presented in
*Corresponding author: Kees Hengeveld: Amsterdam Center for Language and Communication,
University of Amsterdam, Spuistraat 210, 1012 VT Amsterdam, The Netherlands.
E-mail: p.c.hengeveld@uva.nl
Marize Mattos Dall’Aglio Hattnher: Departamento de Estudos Linguísticos e Literários,
Universidade Estadual Paulista, Rua Cristóvão Colombo 2265, 15054-000 São José do Rio Preto,
SP, Brazil. E-mail: marize@ibilce.unesp.br
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
Section 3. On the basis of the resulting classication we specify three typologi-
cal predictions that follow from it in Section 4. These predictions concern the
potential co-occurrence of evidential markers in a single clause and the implica-
tional relations between the existence of evidential subcategories within a lan-
guage. The latter general prediction is formulated from a qualitative perspective
and from a quantitative perspective. Before putting these predictions to the
test,we rst go into a number of methodological issues in Section 5, which pres-
ents the sample of native languages of Brazil used in the current study and pro-
vides criteria for the identication of the four evidential subcategories. Sections
6–8 then give the results for the three predictions. In Section 9 we present our
conclusions.
Functional Discourse Grammar
.Layering
Since the eighties, a number of grammatical theories have incorporated the idea
that grammatical categories are organized in scopal layers. The basic idea may
be illustrated with the following example from Hidatsa:
() Hidatsa
wíra  i ápáariki stao ski
tree it grow  . 
‘The tree must have begun to grow a long time ago.’
(Matthews 1965)
In this example it is clear that the certainty that is being expressed through the
particle ski does not only involve the lexical content of the utterance. It also in-
volves the fact that the event expressed within the utterance took place in the
remote past, as expressed through the particle stao, and that this event is viewed
from its starting point, as expressed through the morpheme ki. Similarly, the re-
This assumption is most prominently present in Role and Reference Grammar (Foley and Van
Valin 1984), Usage-based Grammar (Bybee 1985), Functional (Discourse) Grammar (Hengeveld
1989; Hengeveld and Mackenzie 2008), and Generative Grammar (Pollock 1989; Cinque and
Rizzi 2011). A major dierence between these approaches is that in Usage-based Grammar and
Functional (Discourse) Grammar layers are dened in semantic terms, while in Role and Refer-
ence Grammar and Generative Grammar they are dened in positional terms. For a detailed com-
parison between various approaches to layering see Narrog (2009).
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
mote past morpheme stao not only situates the event expressed lexically in the
remote past, it also does so for the fact that this event is viewed from its starting
point. The scope relations between the tense, mood, and aspect (TMA) markers in
this sentence are thus as indicated in (2):
() certainty (remote past (ingressive (predicate+arguments)))
.Layers
In Functional Discourse Grammar scope relations are dened in terms of dierent
pragmatic and semantic layers. Pragmatic layers together constitute the Interper-
sonal Level in this model, while semantic layers together constitute the Represen-
tational Level.
At the Interpersonal Level scope relations are dened in terms of dierent
pragmatic layers. The ones that are relevant for our argumentation below are,
working inside out, the communicated content, which represents the message
transmitted in an utterance; the illocution, which species the communica-
tive intention of the speaker; and the discourse act, which is the basic unit of
communication.
At the Representational Level scope relations are dened in terms of dierent
semantic layers. Working inside out again, the layers that are relevant for the
argumentation below are the situational concept, which provides the basic char-
acterization of a state-of-aairs; the state-of-aairs, which is the situated real or
hypothesized situation the speaker has in mind; the episode, which is a themat-
ically coherent combination of states-of-aairs that are characterized by unity
orcontinuity of time, location, and participants; and the proposition, which is a
mental construct entertained about an episode.
The layers within each level are hierarchically related and so are the levels
among themselves. These hierarchical relations are indicated in Figure 1.
Note that the correlation between scopal layers and the relative order of TMA (and E: evidenti-
ality) markers present in Example (1) and its English translation only holds under restricted con-
ditions, namely only to the extent that TMAE markers are expressed using the same morphologi-
cal strategy (Boland 2006: 234–249). Thus the prediction holds for e.g., all axal expressions
among themselves, all particles among themselves, all auxiliaries among themselves, all clitics
among themselves, but not for combinations of e.g., axes, auxiliaries, and particles. Since the
grammars available to us do not always allow us to strictly distinguish between axes and clitics
on the one hand, and particles and auxiliaries on the other, we will not take the linear order of
evidential expressions into consideration.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
Figure 1 shows the hierarchical relations between layers and levels, with the
symbol “>” and “” showing the directions in which layers and levels have scope
over one another. Thus, the Interpersonal Level has scope over the Representa-
tional Level, and within each level layers more to the le have scope over layers
more to the right.
.TMA categories
TMA categories are captured in FDG through operators applying at the dierent
layers of the Interpersonal and Representational Level. The position of an opera-
tor thus reveals the scope of the corresponding TMA category. We follow here the
classication of TMA categories oered in Hengeveld (2011) and Hattnher and
Hengeveld (forthc.). Table 1 summarizes this classication.
Tense, mood, and aspect are not unied categories in their application to
these layers of pragmatic and semantic organization, but fall into dierent sub-
categories according to their scope. Aspect is subdivided into two categories,
Fig. 1: Scope relations in FDG
Table 1: TMA categories in Functional Discourse Grammar
Interpersonal
Level
discourse act illocution communicated
content
Mood basic illocution
Representational
Level
propositional
content
episode state-of-
aairs
situational
concept
Aspect event
quantication
phasal aspect
(im)perfectivity
Tense absolute tense relative tense
Mood subjective
epistemic
modality
objective
epistemic
modality
event-oriented
modality
participant-
oriented
modality
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
separating quantitative aspectual distinctions (such as habitual), which quantify
over states-of-aairs as a whole, from qualitative aspectual distinctions (such
as imperfective), which aect the internal temporal organization of a state-of-
aairs. Tense is subdivided into absolute tense distinctions (such as past), which
locate episodes, i.e., series of states-of-aairs, in time with respect to the moment
of speaking, and relative tense distinctions (such as anterior), which locate a
single state-of-aairs in time relative to another one. The widest range of sub-
categories is found in the area of Mood, where we nd basic illocutions (such as
interrogative), which show the speaker’s communicative intention; subjective
epistemic modality distinctions (such as certainty), which indicate the speaker’s
attitude toward a propositional content; objective epistemic modality distinc-
tions (such as possibility), which provide an assessment of the reality status of
aseries of states-of-aairs; event-oriented modality distinctions (such as moral
obligation), which specify the existence of general facilitating conditions, desir-
abilities, and general obligations; and participant-oriented modality distinctions
(such as ability), which express a relation between a participant in a state-of-
aairs and the realization of that state-of-aairs.
In Section 3 Table 1 will be expanded so as to include evidentiality distinc-
tions. The interaction of evidentiality with the illocutionary and tense categories
specied in Table 1 will turn out to be of special importance in establishing evi-
dential subcategories.
.Grammaticalization
A further point that is relevant in relation to our argumentation below concerns
the treatment of grammaticalization in FDG. The current hypothesis within FDG
(Hengeveld 2011) and in other frameworks (see e.g., Roberts and Roussou 2003;
Narrog 2009) is that generally operators start out as lexical elements and in a pro-
cess of grammaticalization may acquire grammatical status at any layer of gram-
matical structure. Once they have acquired grammatical status, they may acquire
further grammatical functions, but only in two directions: (i) by increasing their
scope layer by layer within the same level, (ii) by moving up from the Representa-
tional to the Interpersonal Level. When applying generalization (i) to Figure 1 and
Table 1, a TMA marker expressing, for instance, phasal aspect at the layer of the
situational concept may develop into a marker of relative tense at the layer of the
state-of-aairs, and subsequently become a marker of absolute tense at the layer
of the episode. According to generalization (ii) a marker of, for instance, dubita-
tive modality at the propositional layer of the representational level may develop
into a marker of interrogative illocution at the interpersonal level.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
 Evidentiality in Functional Discourse Grammar
.Introduction
Using the framework sketched in Section 2, we now turn to the classication
ofevidential distinctions based on scope considerations. These considerations
lead us to posit four evidential subcategories: reportativity, inference, deduc-
tion, and event perception. We rst give a general characterization of these
subcategories in Sections 3.23.5. Only aer this rst presentation we will pro-
vide proof for their distinct nature by discussing some grammatical features
that crucially distinguish them in Section 3.6. Aer an intermediate summary
in Section 3.7, we discuss further subdistinctions within each subcategory of
evidentiality in Section 3.8, and compare our classication with existing ones in
Section 3.9.
.Reportativity
The rst subcategory of evidentiality is reportativity. Reportativity distinctions
indicate that the source of the information that the speaker is passing on is an-
other speaker. In terms of the distinctions made above, this means that repor-
tativity operates at the layer of the communicated content at the Interpersonal
Level: the message content contained in a discourse act is characterized as
transmitted rather than originally produced. The high scope of the reportativity
operator is reected in the fact that the report it introduces may contain all
kindsof material related to the original rather than the current speaker. Consider
Example (3):
() I was told that Sheila will probably come.
By taking scope as a crucial classicatory property of evidential distinctions we take a position
that is diametrically opposed to Boye (2010), a paper which “rejects the idea that dierent types
of evidential meanings have dierent scope properties” and “argues that evidential meanings
share scope properties in the sense that they are all conceptually dependent on a “proposition
– i.e., a meaning unit which can be said to have a truth value.” We do not agree with the argu-
ments presented in Boye (2010), but discussing these would lead us too far away from the main
objectives of the current paper. On an empirical basis this paper will show that, when one does
take scope properties into account in dening evidential distinctions, important typological ge-
neralizations can be arrived at.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
In this sentence the propositional attitude expressed by probably can only be in-
terpreted as expressing the subjective evaluation of the original speaker, not the
current one.
In Example (4), from Lakondê, the sux -setaw is used in this reportative
function:
()Lakondê
ta’wḛn  ’teh-’naw ta-’a̰jh-wi-setaw-’tãn
woods path- -walk-1.--
‘Let’s walk to the path in the woods, someone told me.’
(Telles and Wetzels 2006: 240)
.Inference
The second subcategory of evidentiality is inference. We use this term exclusively
for evidential expressions that the speaker uses to indicate that he infers a certain
piece of information on the basis of his/her own existing knowledge. An utter-
ance characterized by an inferential operator thus elaborates on that existing
andstored knowledge rather than reacts to external perceptual stimuli. In terms
of the distinctions made above it operates at the layer of the propositional content
at the Representational Level. This layer deals with mental constructs as repre-
sented in the speaker’s brain. Inference is dierent from the expression of epis-
temic (un)certainty as in the latter case the proposition brought forward is pre-
sented as (un)certain not as a result of an active inferential process, but because
the relevant knowledge is already stored as (un)certain in the mind of the speak-
er. In cases in which languages use the same marker for inference and epistemic
modality, we include the marker in our data analysis.
In Karo, the evidential particle memã is used when the information conveyed is
an inference based on a known pattern of behavior of the subject of the sentence:
() Karo
=ket-t memã
3.=sleep- 
‘I suppose he is sleeping.’
(Gabas 2004: 269)
In the following example from Desano the speaker likewise bases an inference
on his/her knowledge of the habits of the subject of the sentence, not on any
perceived evidence:
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
() Desano
suʔri koe-go ii--bõ pera-ge
clothes wash-. do--3.. port-
‘I guess she is washing clothes at the river landing.’
(Miller 1999: 67)
.Deduction
The third category of evidentiality is deduction. We use this term for the evi-
dential distinctions that are used to indicate that the information the speaker
presents is deduced on the basis of perceptual evidence. In terms of the dis-
tinctions made above deduction operates at the layer of the episode. This
conclusion is warranted by the fact that deduction necessarily involves at
least two related states-of-aairs: the perceived one and the deduced one.
The speaker deduces the occurrence of one state-of-aairs, the deduced one,
on the basis of another state-of-aairs, the perceived one. Note that the out-
put of the deductive process is propositional in nature, and in that sense
deduction resembles inference, but for the localization of an operator the
inputittakes is decisive. Further support for this analysis will be provided in
Section 3.6.
In Tariana, the evidential sux -nihka is used “to refer to something one has
not seen, but which is based on obvious evidence which can be seen” (Aikhen-
vald 2003: 287288). Thus, in (7) the speaker obtained his/her knowledge through
a deduction on the basis of visual evidence.
() Tariana
tʃinuniwhã-nihka di-na
dog 3...bite-.. 3..-
‘The dog bit him (I can see obvious signs).’
(Aikhenvald 2003: 288)
Deduction is frequently based on visual evidence, but not exclusively. In (8) and
(9), the speaker bases his/her deduction on sounds and smells, respectively:
() Yuhup
̰ɉìdə̆h    ̃ɉábmá  ̃hó
3 dance
‘They are dancing.’ (as I deduce from the noise).
(Ospina Bozzi 2002: 183)
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
() Sabanê
kieylali-k kan-n-tika hala-n-dana
peccary- die--. stink--..
‘The peccary died; (because) it stinks.’
(Araújo 2004: 143)
The fact that deduction is always based on perceptual evidence is easily seen in
Sabanê, in which the marker of deduction appears in sentences that are obliga-
torily accompanied by another sentence which contains a sensory evidential, as
in (9).
.Event perception
The fourth category of evidentiality is event perception. By means of evidential
expressions of this type the speaker indicates whether or not he witnessed the
event described in his utterance directly. With “directly” we mean that the speak-
er was at the scene and through one of the senses perceived the occurrence of a
state-of-aairs. Perception is thus involved in both deduction and event percep-
tion. The crucial dierence is that in the case of deduction the state-of-aairs that
the utterance is about is not perceived through one of the senses (though another
one is, the one that forms the basis for the deduction), while in the case of event
perception the state-of-aairs that the utterance is about is perceived through one
of the senses. In terms of the distinctions made above event perception operates
at the layer of the state-of-aairs, as it is this state-of-aairs that is directly
perceived.
The following examples from Tuyuca illustrate visual event perception and
non-visual event perception respectively:
() Tuyuca
díiga apé-wi
soccer play-.
‘He played soccer.’ (I saw him play.)
(Barnes 1984: 257)
() Tuyuca
mũtúru  bɨsí-tɨ
motor roar-.
‘The motor roared.’
(Barnes 1984: 260)
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
Event perception also includes elements that express that an event was not per-
ceived directly. An example of such an element, in contrast with an element
expressing direct perception, is given in (12):
() Jarawara
Wero kisa-me-no
name() get.down--...
ka-me-hiri-ka
in.motion--...-.
‘Wero got down from his hammock (which I didn’t see), and went out (which
I did see).’
(Dixon 2004: 204)
Note that, in contrast with e.g., Willett (1988) and Aikhenvald (2004) we con-
siderthe non-witnessed category to be an instantiation of event perception, not
an expression of indirect evidentiality. Deduction, inference and reportativity
all imply the absence of perception, but these evidential interpretations can
all be derived from the basic meaning of the non-witnessed category of event
perception.
. Distinguishing features of the four evidential
subcategories
In the preceding sections the arguments for distinguishing between the four
subcategories of evidentiality and their association with certain layers within the
FDG model are primarily based on the semantics of evidentials in combination
with the semantics of layers. Evidence for the relevance and separate status of
these subcategories is, however, also supported by further grammatical evidence.
This evidence has to do with the combinability of the four evidential subcatego-
ries with basic illocutions and with tense.
The combinability of evidential subcategories with various basic illocutions
allows us to separate reportativity from the other three types of evidentiality.
Inmost cases evidentiality is limited to declarative and, with certain restrictions,
interrogative sentences. Declaratives and interrogatives are related in the sense
that both are concerned with the transmission of information. They contrast with
basic illocutions that are concerned with inuencing behaviour, such as impera-
tives and hortatives. Reportativity allows the combination with both of these
types of basic illocution. This is a result of the fact that in principle any type of
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
sentence can be reported, as is also evident from the existence of, for example,
relayed imperatives in certain languages. The following examples are from Hup
(Epps 2008: 654):
() Hup
́m-̃́y=mah
fear-=
‘(He’s) scared, he says.’
(Epps 2008: 655)
() Hup
-n’ǐ h=mah?
-=
‘What did he say?’
(Epps 2008: 655)
() Hup
nǽn=mah!
come=
‘Come here, he said!’
(Epps 2008: 656)
The reportative combines with a declarative in (13), with an interrogative in (14),
and with an imperative in (15). The reportative is the only evidential in Hup that
may combine with the imperative (Epps 2008: 656). This is also true of the repor-
tative in Tariana, as observed by Aikhenvald (2003: 322). An example of this com-
bination is given in (16):
() Tariana
pi-a pi-ñha-pida
2.-go 2.-eat-.
‘Go and eat.’ (on the order of someone else)
(Aikhenvald 2003: 376)
Similarly, in the reportative examples from Lakondê given above in (4) the repor-
tative is used in a hortative sentence, again showing the compatibility of reporta-
tivity with illocutions of the behavioural type.
The other three subtypes of evidentiality can be distinguished from one an-
other in terms of their interaction with the categories of absolute and relative
tense. Since in our sample languages in many cases evidentiality and tense are
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
expressed in portmanteau morphemes, we cannot use examples from these lan-
guages to illustrate the interaction between evidentiality and tense. Instead, we
will use examples in which lexical complement-taking verbs express the various
evidential values so that we can show the interaction between the various mean-
ing components. By applying this strategy we by no means want to suggest that
grammaticalized evidentiality and the lexical expression of evidential values
should be taken as forming a single system. It does, however, allow us to tease
the various evidential and temporal meanings apart. Compare the following
examples:
() Inference
I infer that he is cooking.
() Deduction
I smell that he is cooking.
() Event perception
I see him cooking.
These examples are similar in that the temporal specication of the main clause
is identical to that of the subordinate clause in all three cases.
A rst dierence between the three evidential scenarios illustrated in (17)–
(19) shows up in (20)–(22):
() Inference
I infer that he has been cooking.
() Deduction
I smell that he has been cooking.
() Event perception
*I see him having been cooking.
A dierence between inference and deduction on the one hand and event per-
ception on the other is that relative tense modications are allowed within the
scope of the former but disallowed within the scope of the latter. The ungrammat-
This is a case of deduction, as from the smell one cannot identify who is doing the cooking.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
icality of (22) is a consequence of the fact that direct perception necessarily im-
plies simultaneity. This sets o event perception from the remaining two eviden-
tial categories.
A dierence between inference and deduction shows up in (23)–(24):
() Inference
I infer that he had been cooking.
() Deduction
*I smell that he had been cooking.
A dierence between inference on the one hand and deduction on the other is
that absolute tense modications are allowed within the scope of the former but
disallowed within the scope of the latter. The ungrammaticality of (24) follows
from the fact that in order for one state-of-aairs to provide evidence for the oc-
currence of another, there has to be a temporal connection between the temporal
reference point of the state-of-aairs providing the evidence and that of the de-
duced state-of-aairs. This does not mean that the deduced state-of-aairs has to
have occurred before the state-of-aairs providing evidence: the inverse temporal
order is possible too. Suppose someone is expecting someone else in a hallway
where elevators are situated and he/she is aware of the fact that this person has
to come from the second oor to where he/she is. The lights indicating where the
elevator is may then provide evidence for him/her to predict that the person he/
she is expecting is on his way, as in:
()I can see that he is going to arrive.
The four subcategories of evidentiality are thus dierent in their behaviour as
regards their co-occurrence with basic illocution and absolute and relative tense,
as indicated in Table 2.
Table 2: Distinctive features of evidential subcategories
Evidential Criterion
Subcategory
Combines with
behavioural illocutions
Takes absolute tense
within its scope
Takes relative tense
within its scope
Reportativity + + +
Inference − + +
Deduction − +
Event Perception
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
. Intermediate summary
Table 1 may now be expanded to include evidential subcategories as in Table 3.
Table 3 shows how evidential subcategories are situated in relation to other
TMA categories. It specically shows that only reportativity operates at the Inter-
personal Level, where it interacts with basic illocution, and that inference and
deduction are situated at layers where they dominate tense categories at lower
layers. Inference dominates absolute and relative tense, deduction only domi-
nates relative tense, and event perception dominates neither. The classication
arrived at thus correctly reects the interactions between evidentiality on the one
hand, and basic illocution and tense on the other, as argued in Section 3.6.
.Evidential sub-subcategories
The fact that we can identify four subcategories of evidentiality in a language
does not necessarily mean that this language may only have four evidential
markers. Within each subcategory further distinctions are possible. In Ma-
maindê,for instance, there are two markers of event perception: one to indicate
that the speaker perceived the occurrence of a state-of-aairs through the visual
sense (26) and another one to indicate that the speaker perceived the state-of-
aairs through one of the non-visual senses (27):
Table 3: TMA and evidential categories in Functional Discourse Grammar
Interpersonal
Level
discourse act illocution communicated
content
Mood basic illocution
Evidentiality reportativity
Representational
Level
propositional
content
episode state-of-aairs situational
concept
Aspect event
quantication
phasal aspect
(im)perfectivity
Tense absolute tense relative tense
Evidentiality Inference deduction event perception
Mood subjective
epistemic
modality
objective
epistemic
modality
event-oriented
modality
participant-
oriented
modality
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
() Mamaindê
ta-tukwinʔni-tu na-ʔaik-tu
.1.-father.in.law- .3.-eld-
tau-lat ʰa-wa
chop-.3..--
‘My father-in-law is clearing his eld.’ (and I know this because I just came
from his eld and I saw him working)
(Eberhard 2009: 474)
() Mamaindê
ta-tukwinʔni-tu ʔaik-tu
.1.-father.in.law- .3.-eld-
tau-ø-nha-wa
chop-.3-..-
‘My father-in-law is clearing his eld.’ (and I know this because I just passed
near his eld and heard him chopping)
(Eberhard 2009: 475)
In both cases, the evidential concerns the direct perception of a state-of-aairs by
the speaker.
In Mamaindê there are two dierent reportative suxes too, one expressing
that the original utterance was produced by a second hand source (-satau), the
other that it was produced by a third hand source (-sĩ̃n):
() Mamaindê
waʔnĩn-soʔka janãn-tu
shaman-. jaguar-
sun-satau-le-ø-hĩn-wa
kill--.-.3-..-
‘The shaman killed a jaguar (yesterday).’ (and I know this because someone
told me)
(Eberhard 2009: 478)
() Mamaindê
ta-tukwinʔni-tu ʔaik-tu tau-sĩn-ø-nha-wa
1-father.in.law- eld- chop--.3-..-
‘My father-in-law is clearing his eld.’ (and I know this because someone
said they were told that it was so)
(Eberhard 2009: 480)
In both cases, the evidential concerns the report of a communicated content.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
An overview of the subdistinctions made in the languages of our sample
within each of the four evidential subcategories is provided in Section 7.
. Comparison with other classications
The four evidential subcategories we discuss above have been recognized under
dierent names in the literature on evidentiality, and we do not claim any origi-
nality in this sense. What is dierent in our approach is that we claim that these
four subcategories are all primary subcategories of evidentiality, as each has
scope over a dierent layer of grammatical structure. Other authors make dif-
ferent subgroupings, and we will review these briey in this section. Table 4 pro-
vides a systematic comparison.
Willett (1988) makes a primary distinction between direct and indirect evi-
dence. Within the category of indirect evidence he further distinguishes between
reported evidence (the speaker received the information from another speaker)
and inferred evidence (the speaker inferred the situation from its results or
through logical reasoning). Within the category of inferred evidence Willett thus
distinguishes between our Deduction category (inference from results) and our
Inference category (inference through reasoning).
De Haan (1998) too argues that, although evidential systems may dier from
language to language, the opposition between direct and indirect evidence is the
basis of all of them. He makes further subdistinctions, though not between our
Deduction and Inference.
The six semantic values attested by Aikhenvald (2004) in her crosslinguistic
analysis of grammatical evidentiality are visual, non-visual sensory, inference,
assumption, hearsay, and quotative, which combine in dierent ways in eviden-
tial systems. Her categories can be directly mapped onto ours, but Aikhenvald
does not posit a priori further groupings, as we do.
Although Plungian (2010: 29–30) considers the opposition between direct
and indirect access to information the hierarchically and typologically most
important one, he also suggests that the opposition between personal and
non- personal access to information may play an important role in the iden-
tication of evidential subtypes. Besides the predictable combination of direct
and personal access or indirect and non-personal access, the combination
of indirect and personal access to a situation is possible too: “In this case the
speakers obtained knowledge of a situation themselves, without other persons
being involved, but the knowledge of this situation has not been obtained
in a direct way since the speakers did not observe the situation directly.” The
two evidential subtypes identied as characterized by indirect and personal
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
access are inferential (the speaker draws a logical conclusion on the basis of
observed results, our Deduction) and presumptive (the speaker draws a con-
clusion based on his/her knowl edge of the world, our Inference). Plungian
furthermore within the direct group adds the category of participatory evi-
dentiality, which refers to evidence obtained through participation in a state of
aairs.
Finally, San Roque and Loughnane (2012) basically take over Willett’s (1988)
classication, but add Plungian’s (2010) category of participatory evidentiality.
A systematic comparison between the evidential subcategories proposed by
these authors and the ones resulting from our approach is shown in Table 4. Table
4 shows that the major dierence between our approach and the other ones
discussed here is that the main split in our approach is between reportativity
andallother types of evidentiality, following from the major division between
Table 4: Comparison of classications of evidential categories
Source Classication of evidential categories
This article Representational Interpersonal
Event Perception Deduction Inference Reportativity
Willett
()
Direct Indirect
Attested Inferring Reported
Visual Audi-
tory
Other Results Reasoning Second-
hand
Third-
hand
de Haan
()
Direct Indirect
Visual Audi-
tory
Other Inferential Quotative
Plungian
()
Direct Access Indirect Access
Personal Personal Non-personal
Partici-
patory
Visual Non-visual Inferential Presump-
tive
Reportative
San Roque
and
Loughnane
()
Direct Indirect
Partici-
patory
Visual Sensory Inferring Reported
Results Reasoning
Aikhenvald
()
Visual Sensory Inference Assump-
tion
Hearsay Quota-
tive
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
interpersonal and representational categories in FDG. As we show in Sections 6
and 7, our separation of reportativity from other evidential subcategories allows
us to formulate an evidential hierarchy, presented in the next section, in a
straightforward manner.
Predictions
On the basis of the theory outlined above, we now come to a number of predic-
tions as regards the distribution and expression of evidential operators.
The co-occurrence of evidential subcategories: A rst and rather straightfor-
ward prediction that follows from our approach to evidentiality, in which the
four evidential subcategories proposed are located at four dierent layers of
underlying pragmatic and semantic structure, is that markers of these four
dierent subcategories will be allowed to co-occur in a single clause.
The existence of evidential subcategories (qualitative): Our second predic-
tion is that there will be an implicational relationship between eviden-
tialmeanings present in a language according to the following evidentiality
hierarchy:
()event perception  deduction  inference
As explained in Section 2.4, according to FDG theory grammatical elements
may acquire new meanings (i) by increasing their scope layer by layer with-
inthe same level, (ii) by moving up from the Representational to the Inter-
personal Level. According to (i) an evidential expressing, for instance, visual
event perception at the layer of the state-of-aairs may come to express de-
duction on the basis of visual evidence at the layer of the episode. Only aer
that may it acquire the function ofexpressing inference at the layer of the
proposition. According to (ii) any evidential from the Representational Level
may come to express reportativity. On the basis of (i) one may expect that
evidentiality at higher layers can only exist by virtue of its existence at lower
layers, hence the hierarchy in (30). On the basis of (ii) one expects that repor-
tativity does not play a role in this implicational relationship, hence its ab-
sence from (30). Note with respect to (30) that in principle grammatical ele-
ments at higher layers might also develop directly out of lexical elements.
However, we expect that this will only occur when the more basic categories
at lower layers are present in a language. Furthermore, in principle a lower
layer element may disappear when it develops into a higher layer one. Again,
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
we expect this will only happen when an alternative expression for the more
basic lower category is available.
The existence of evidential subcategories (quantitative): Following the same
type of reasoning we predict that the implicational hierarchy in (30) will also
manifest itself in a quantitative sense, such that the number of distinctions
made within each of the subcategories of evidentiality within a single lan-
guage will decrease from le to right in (30). This prediction follows from
theexpectation that new distinctions typically arise at the lowest evidential
layer, are closer to their lexical origin, and will only in some cases make it
tohigher layers. Note that for methodological reasons we only count overt
grammatical markers and not the potential zero marking of an evidential
subcategory, as in many of the languages of the sample evidentiality is an
optional category.
In Sections 6–8 these three predictions are tested against data from a sample con-
sisting of native languages of Brazil. In Section 5 we rst present this sample and
explain how the data from this sample was processed.
Methodological issues
.The sample
This study forms part of a larger research enterprise that aims to establish a
comprehensive typology of the native languages of Brazil. Brazil is home to a
large variety of languages, virtually all of which are in danger of extinction. Lewis
(2009) lists 226 extant and extinct spoken native languages in 21 major groups for
Brazil. Though many new descriptions have become available over the last ten
years, the majority of these languages have hardly been documented, which
makes it dicult to draw up a representative sample. For this reason, all relevant
languages for which we had access to a grammatical description at the time of the
research are included in the sample. It is important to mention that our classica-
tion of the sample languages is based on the descriptions we have access to and
that these may not cover all details of the evidential systems of the languages
For some earlier results see Hengeveld et al. (2007) and Hengeveld et al. (2012).
Lewis (2009) also lists 2 sign languages, 2 creole languages, 1 mixed language, and 5 Indo-
European languages for Brazil, all o which are excluded from the present research. Unclassied
languages are treated as one group in Table 5.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
Table 5: Sample languages
Language family # Lgs Sample languages
ARAUAN Jamamadí, Jarawara
ARAWAKAN  Baré, Palikur, Parecis, Tariana, Terena, Wapixana
ARUTANI-SAPE  –
CARIB  Apalaí, Kuikuro, Macushi, Tiriyó, Waiwai
CHAPAKURA-WANHAM  Moré
KATUKINAN Katukina-Kanamari
MACRO-GE 
BORORO  –
BOTOCUDO  –
FULNIÔ Fulniô
GE-KAINGANG  Apinayé, Parkatêjê, Pykobje, Xavante
GUATÓ  Guató
KAMAKAN  –
KARAJA  –
MAXAKALI  Maxacalí
OPAYE  –
OTI  –
PURI  –
RIKBATSA Rikbáktsa
YABUTI
MAKU Dâw, Hup, Nadëb, Yuhup
MATACO-GUAICURU Kadiwéu
MURA Pirahã
NAMBIQUARAN Lakondê, Mamaindê, Sabanê, Nambikuára
PANKARARÚ
PANOAN  Amahuaca, Huariapano, Katukina, Kaxinawá, Matses,
Shanenawa, Yaminahua
TICUNA
TRUMAÍ Trumaí
TUCANOAN  Carapana, Cubeo, Desano, Tuyuca, Wanano, Ye’pâ-masa
TUPI 
ARIKEM Karitiana
AWETI Aweti
MAWE-SATERE
MONDE Surui
MUNDURUKU Munduruku
PURUBORA
RAMARAMA Karo
Yuhup is spoken in both Brazil and Colombia. The grammar we use (Ospina Bozzi 2002) de-
scribes the Colombian variety.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
concerned. The total number of these languages is 64, which gives a coverage of
28% of the native languages of Brazil. The languages distribute across the afore-
mentioned groups in the way indicated in Table 5, in which the rst column gives
the name of the language family concerned, the second column the number of
languages in this family spoken in Brazil, and the third one the sample languages
from this family. Only languages with at least one grammaticalized evidential
cate gory, 34 out of the 63 languages investigated, are relevant to the predictions
investigated in the present article. Languages without such a category are under-
lined in Table 5 and will not gure in some of the overviews presented later in
thispaper. Note that we give the languages names as they are used by the au-
thorsof the principal reference grammars on which our research is based. For
anoverview of the language descriptions that were consulted for this study see
Appendix B.
. The identication of evidential subcategories
The information about evidential systems of the sample languages was collected
from existing reference grammars. An overview of these sources is given in Ap-
pendix B. The grammars dier not only in extension and depth but also in the
descriptive labels that are used for evidential subcategories. In order to com-
pare these languages, all evidential subcategories described in the grammars
arereclassied in terms of the labels presented above in the context of the FDG
In establishing the sample the inventory presented in Torres Sánchez (2013) has been very
useful to us.
Table 5 (cont.)
Language family # Lgs Sample languages
TUPARI
TUPI-GUARANI  Guajá, Guajajara, Guarani-Mbyá, Kamaiurá, Kokama-
Kokamilla, Nheengatú, Parintintín, Urubu-Kaapor
YURÚNA
TUXÁ
UNCLASSIFIED  Kanoê, Kwaza
WITOTOAN
YANOMAM Sanuma, Yanomam
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
approach to evidentiality. In the case of the 34 languages with evidentiality re-
tained for the present research, the semantic information oered by the gram-
mars is sucient to carry out such a re-classication with a reasonable degree of
condence.
The easiest subtype to be identied is Reportativity, as language descrip-
tionsrefer to this category in a consistent way. The names used to identify this
category vary, but their denitions are the same, as can be seen in the following
descriptions:
The mediative evidential particle indicates that the content was attested by someone else or
by hearsay. (Guajá, Magalhães 2007: 84 [our translation])
The reportative sux -kia indicates that the information is second-hand. It can be trans-
lated by it is said that … or they say that … (Yaminahua, Faust 2002: 30 [our translation])
[The hearsay particle tsile] is used when the speaker wants to report a fact that s/he did not
observe and for which there is no direct evidence. The speaker knows the fact because
somebody else told it to her/him. The particle is found very oen in mythical narratives and
it is translated by the consultants as ‘they say, people say that …’ (Trumaí, Guirardello 1999:
225)
Evidentials of Event Perception are also consistently described and their identi-
cation generally does not oer any special diculty. Some languages have dier-
ent ways to indicate that the speaker had direct access to the information he/she
conveys, expressing whether the event was perceived through the visual or a non-
visual sense, as in Tuyuca:
[…] visual evidentials are used to describe states or events that the speaker saw or is seeing,
including those in which he himself is the actor […] nonvisual evidentials are used to report
how someone, something or some event smelled, sounded, tasted, or felt (smells, sounds,
tastes, or feels). (Tuyuca, Barnes, 1984: 259)
Other languages only express that the event was directly perceived by the speak-
er, without distinguishing between senses. Evidentials with this meaning receive
labels such as ‘observation’ (Pirahã), ‘attested’ (Jamamadi), ‘witness’ (Sanuma)
or ‘experiential’ (Matses):
The term ‘Experiential’ should be dened carefully here to distinguish it from ‘Inferential’
(next section). The essential condition is that the speaker witnesses the event (using any of
the ve senses) as the event happens. A denition could be as follows: experiential refers to
a situation where the speaker detects the occurrence of an event at the time that it transpires
(or a state at the time that it holds true). The primary (i.e., optimal or most direct) way of
detecting most events is by visual contact, but not always. (Matses, Fleck 2003: 402)
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
The identication of Deduction and Inference is harder for dierent reasons.
First, due to the similarities in the meanings of the words inference and de-
duction, one and the same evidential meaning is sometimes identied with
either of these labels. This is for instance the case in the descriptions of
Wanano and Ye’pâ-masa. These two languages have grammatical ways to
indicatethat the information the speaker conveys is deduced by him/her based
on perceived evidence, but still these evidentials are classied with dierent
names:
Inference markers are used in utterances in which the speaker is presenting a conclusion
about an event or state based on directly perceived results, inferring what happened based
on the current evidence (Wanano, Stenzel 2004: 357)
The deductive modality forms are used when the speaker did not see or perceive the verbal
situation but has proof (traces) that this situation happened. (Ye’pâ-masa, Ramírez 1997: 137
[our translation])
In both of these cases we classify the evidential as expressing Deduction, since
inboth cases the speaker deduces the occurrence of an event through the per-
ception of existing evidence. We classify an evidential as expressing Inference
only in those cases in which the speaker comes to his/her conclusion on the
basisof his/her existing store of knowledge. Examples of denitions of inference
markers under various labels are the following:
Assertion suxes are used to code statements in which the speaker’s assessment of a situa-
tion is based not on any specic currently accessible outside evidence, but on internal or
internalized evidence. This evidence can be founded either on the speaker’s own previous
experience, upon which s/he can make reasoned suppositions, or on his/her cultural, his-
torical, or physical knowledge of the world, upon which s/he can make assertions of fact.
(Wanano, Stenzel 2004: 359)
The assumed evidential tells the hearer that the speaker has not seen or is not seeing the
event, but supposes that an event has occurred or is occurring based on his knowledge
ofthe habits of the persons involved, what they indicated they were going to do, or on his
general knowledge of how things work. (Desano, Miller 1999: 66)
A strong supposition is indicated by the marker kite. This supposition usually refers to
future time, but can refer to the present, both present and future at once, and the past.
(Sanuma, Borgman 1990: 172)
On the basis of these and other denitions, the re-classication is possible.
In Table 6 we provide an overview of the correspondences between our
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
Table 6: Terminology used in reference grammars in relation to our terminology
Representational Interpersonal
Event
Perception
Deduction Inference Reportative
Apalaí Eyewitness
evaluation
Deduction or
assumption
Evidently
Carapana Denido Evidente Reportativo
Cubeo Witnessed Assumed Reportative
Dâw Reportativo
Desano Visual Inferred Assumed Hearsay
Guajá Atestada Mediativa
Huariapano Evidencial direto Reportativo
Hup Nonvisual Inferred Inferred  Reportative
Jamamadí Testemunha
ocular
Vericada Suposição Relatada por testemunha
ocular, Relatada por
testemunha não-ocular
Jarawara Eyewitness,
Non-eyewitness
Reported
Kamaiurá Evidência direta
visual, Evidência
direta prévia
Inferencial/
dubitativa
Opinativa Reportativa
Karitiana Hearsay
Karo Visual Inference from
evidence
Inference from
expectation
Hearsay
Kokama-
Kokamilla
Reported
Lakondê Visual,
Non-visual
Sensory
evidence
Hearsay, Quotative
Mamaindê Visual,
Non-visual
Inferred General
knowledge
Reported Secondhand,
Reported Third hand
Matses Direct
experience
Inferential Conjecture
Nambikwara Observação
individual,
Obsevação
coletiva
Dedução
individual,
Dedução
coletiva
Narrativa individual,
Narrative coletiva
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
re- classication and the original terms used by the authors of the grammars
thatformed the basis of our research.
In glosses, the four evidential subcategories appear as /.
for markers of Event Perception and the absence thereof,  for markers of
Deduction,  for markers of Inference, and  for markers of Reportativity.
Further subdistinctions are made where relevant.
Table 6: (cont.)
Representational Interpersonal
Event
Perception
Deduction Inference Reportative
Nheengatú Reportativo
Parkatêjê Validacional Inferencial Reportativo
Pirahã Observation Deduction Hearsay
Sabanê Sensory
evidentiality
Inferred Inferred
neutral
Hearsay
Sanuma Witnessed Veried Supposition Veried
Surui Hearsay
Tariana Visual,
Nonvisual
Inferred
specic
Inferred
generic
Reported
Trumai Hearsay
Tuyuca Visual,
Nonvisual
Apparent Assumed Secondhand
Urubu-Kaapor Hearsay
Waiwai First person
responsability
Evidently Third person
responsability
Wanano Visual,
Non-visual
Inference Assertion Hearsay quotative,
Hearsay diuse
Yaminahua Testimonial Reportativo
Yanomam Testimonial,
Non-testimonial
Déductif Citatif
Ye-pâ-masa Vista, Sentida Dedutiva Reportativa
Yuhup Inferentiel  Inferentiel ,
Inferentiel 
Rapportatif
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
Co-occurrence of evidential markers
The rst prediction we formulated in Section 4 concerns the co-occurrence of
evidential subcategories. If it is true that evidentiality is not one category
but actually covers four dierent subcategories applying at dierent layers of
grammatical structure, we expect it to be possible for two or more evidential
expressions from dierent subcategories, i.e., dierent in terms of their scope
properties, to co-occur in a single expression. Before going to the actual data, let
us rst consider the logical possibilities of combining evidential meanings. Note
again that in the following examples we paraphrase the meaning of evidentials
using English lexical verbs. By doing so we by no means want to suggest that
gram matical and lexical expressions of evidentiality form a single system. Our
only purpose is to explicitly separate the dierent evidential meanings, some-
thing we can not do in the same systematic way using the data available to us
from the sample languages.
The maximal combination of four dierent evidential expressions, one from
each evidential subcategory, is paraphrased in (31):
() I hear ( from A) that A inferred on the basis of his existing knowledge that B
deduced from visual evidence that C had been smoking, something that B did
not witness directly.
Of course, such a sentence overloaded with evidentiality distinctions is unnatu-
ral for various reasons. The point here is that it is semantically possible. In order
to show this we will look at the combinations of evidentiality distinctions in all
possible pairs of two.
Reportativity+Inference
() I hear ( from A) that A inferred on the basis of his existing knowledge that C
had been smoking.
Reportativity+Deduction
() I hear ( from A) that A deduced from visual evidence that C had been smoking.
Note that the co-occurrences we list may have either a scoped reading or a concord (see e.g.,
Geurts and Huitink 2006) reading. Our material does not always allow us to tell these two apart.
However, combinations of grammatical evidential markers in a concord relations are just as in-
dicative of the relevance of the dierent evidential subcategories as the scoped combination are:
in both cases the relevant slots for the evidential markers should be available.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
Reportativity+Event Perception
() I hear that from A that C has been smoking, something that A did not witness
directly.
The combinations that involve reportativity are straightforward: everything that
has been communicated can be reported, which means that if an original utter-
ance may contain an evidential marker, the reported utterance will contain that
same marker. Note that in these cases the evidential markers that are within
thescope of the reportativity marker are attributed to the original source. When
an utterance containing a marker of reportativity is reported a special situation
arises. As shown above, Mamainde has a special marker for such a third hand
report, next to its marker for second hand reports.
Inference+Deduction
() I infer on the basis of my existing knowledge that B deduced from visual evi-
dence that C had been smoking.
The combination of inferentiality and deduction is less straightforward. But
imagine I know B very well and that nothing irritates him more than the fact that
C is smoking now and then. I also know that C was just smoking a cigarette and
le the cigarette butt in the ashtray that was empty when B le the house. When
I see B getting mad at C on his return I could say (35). It is clear that one has to
setup complex scenarios to get an appropriate context for this combination of
evidential meanings.
Inference+Event Perception
() I infer on the basis of my existing knowledge that C has been smoking, some-
thing that I did not witness directly.
The combination of inferentiality and event-perception is straightforward again.
When I infer something about a certain state-of-aairs, it follows that I did not
witness that state-of-aairs directly, or I would not use an inferential marker.
Inferentiality thus necessarily implies the absence of visual evidence.
Our prediction thus is that in languages that have markers for more than
one evidential subcategory, members of these subcategories may co-occur in a
single clause. Given the unlikeliness of a speaker of a language wanting to com-
bine many dierent evidential markers in a single clause, as in the improbable
yet grammatical Example (31), we have only found combinations of evidential
markers in the ve pairs illustrated in (32)–(36). Table 7 shows all the attested
co-occurrences.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
As expected, Reportativity co-occurs with all other evidentials, since they
pertain to dierent levels. The following examples illustrate the three possible
combinations.
Reportative+Inference
() Hup
yúp hɔ́tʔah=mah hɨd ye--ip=b’a y h
that other.side= 3. enter--=-
‘There on the other side of it (someone said) they apparently got in again.’
(Epps 2008: 660)
The marker - in Hup tends to be used, in contrast with the marker =cud illus-
trated in (38) below, when the emphasis is on the actual act of inferring, and is
preferred when there is no actual evidence available.
Reportative+Deduction
() Hup
hup pã̌=cud=mah
person .==
‘There was apparently nobody there, it’s said.’
(Epps 2008: 658)
In Hup, the evidential cud is used to designate a deduction based on tangible
proof. This proof is oen, although not necessarily, visual evidence, as in (38).
This deduction is part of the original message that is being reported in (38).
Table 7: Co-occurrence of evidentials
Evidentiality
Language
Event Perception Deduction Inference Reportativity
Hup
Hup
Huariapano, Hup, Jarawara,
Mamaindê, Sabanê
Karo
Wanano
Hup, Sabanê, Wanano
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
Reportative+Event Perception
() Sabanê
wayulupi.maysili-k kan-n-tiaka-dana
cat.younglings- die---
‘Somebody said that the kitten died.’
(Araújo 2004: 154)
In (39), the message being reported concerns a state-of-aairs perceived by the
original speaker.
As shown above, the combination of Inference and Deduction requires very
specic scenarios. We have found only one example of this combination.
Inference+Deduction
() Karo
péŋ aʔ=wĩ-n aketmemã
White.man 3.=kill-  
‘The white man must have supposedly killed it/him.’
(Gabas 1999: 277)
It should be noted, however, that Gabas (1999: 277) remarks the following
aboutthis and other combinations: “In the interview, the consultant said the se-
quences were utterly possible. But I do not have actual examples of their occur-
rences in natural texts/conversations. I am also not certain about their precise
translations.”
Inference+Event Perception
The following example illustrates the combination of inference and event
perception.
() Wanano
bora-~su-ka wa’a-ro koa-ta-a
fall.down-- go- .-come-.
‘He fell right down.’
(Stenzel 2004: 103)
Deduction+Event Perception
The deduction of an event is always based on some evidence available to the
speaker and its expression may be aected by the way he/she accessed this
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
evidence. In the following example from Wanano, the deduction is based on visu-
al evidence. As Stenzel (2004: 358) arms, in (42) the speaker is examining a set
ofbaskets that had been stored for a long period of time. One of the baskets is
deformed, pushed in on one side, prompting her to comment:
() Wanano
a’yootipa-wa’a-ri hi-ra
Oh! be.at-become-. -...1
‘Oh! This one’s (been) attened.’
(Stenzel 2004: 358)
In Sabanê, as already shown in Section 3.4, the co-occurrence of deduction and
event perception is indeed a rule, since the deduction evidential appears only in
sentences preceding another evidential sentence which species the perceived
evidence, as in (43):
() Sabanê
kieylali-k kan-n-tika hala-n-danal
Peccary- die--. stink--..
‘The peccary died; (because) it stinks.’
(Araújo 2004: 143)
The examples presented here show that indeed every possible pair of evidential
subcategories is attested in languages of the sample. We should note that the
combinability of evidentials is hardly ever discussed explicitly in the grammars
that form the basis for this study, so that a systematic comparison of the possibil-
ities is impossible.
 Implicational relations between evidential
meanings (qualitative)
.Introduction
In this section we present the results for our second prediction (Section 7.2), dis-
cuss the consequences of these results for the classication of evidential systems
(Section 7.3), and comment on the areal and genealogical distribution of these
systems (Section 7.4).
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
. The existence of evidential subcategories
The second prediction we make in Section 4, based on considerations having to
do with possible grammaticalization paths, is that there will be an implicational
relationship between evidential meanings present in a language according to the
hierarchy in (30), repeated here:
() event perception  deduction  inference
We also predict that languages may or may not combine any of the resulting sys-
tems with a reportative evidential. The data needed to test this prediction are
given in Table 8.
Table 8 shows that our prediction is fully conrmed. There are no languages
with an evidential subcategory of Inference that do not also have the evidential
subcategories of Deduction and Event Perception; and there are no languages
with an evidential subcategory of Deduction that do not also have the eviden-
tialsubcategory of Event Perception. And if a language has only one evidential
marker from the three in (30), it is one expressing Event Perception. These three
possibilities may or may not be combined with an evidential marker of Reporta-
tivity, which is almost omnipresent, but lacks in one language otherwise rich
inevidentiality. Reportativity may also be the only evidential marker of a lan-
guage. Its high frequency is in accordance with the results of Aikhenvald’s (2003:
31) typological study.
The systems of Parkatêjê, Yuhup, and Sanuma deserve special attention.
InParkatêjê the evidential particle mə̃r may be used to indicate that the speaker
has arrived at a certain conclusion through Deduction on the basis of sensory
evidence, or through Inference based on existing knowledge. These two mean-
ings are located at two contiguous layers of evidential marking, so that the
overlap in meaning is in accordance with our prediction, which is based on
contiguity.
A variant of this situation manifests itself in Yuhup: in this language the
marker indicates that an event is perceived directly through the auditive
channel, or that its occurrence is deduced on the basis of auditive information.
Again, these two meanings are contiguous on the evidential hierarchy presented
in (30).
A last case of contiguity is exemplied by Sanuma. In this language the
marker noa/no may be used both for Deduction and for Reportativity. Though
these two meanings, due to the restrictions of a two-dimensional medium, are
notrepresented contiguously in Table 8, they are contiguous in the theoretical
framework we are using here. As indicated in Section 2.4, in the course of a
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
Table 8: Evidential subcategories in the languages of the sample (* = same marker)
Level Representational Interpersonal
Evidential Event perception Deduction Inference Reportative
Desano + + + +
Hup + + + +
Jamamadí + + + +
Kamaiurá + + + +
Karo + + + +
Mamaindê + + + +
Parkatêjê + +* +* +
Sabanê + + + +
Sanuma + +* + +*
Tariana + + + +
Tuyuca + + + +
Wanano + + + +
Apalaí + + +
Matses + + +
Carapana + + +
Cubeo + + +
Lakondê + + +
Nambikwara + + +
Pirahã + + +
Waiwai + + +
Yanomam + + +
Ye-pâ-masa + + +
Yuhup +* +* +
Guajá + +
Huariapano + +
Jarawara + +
Yaminahua + +
Dâw +
Karitiana (Panoan) +
Kokama-Kokamilla +
Nheengatú +
Surui +
Trumai +
Urubu-Kaapor +
There is a further dedicated marker of deduction.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
grammaticalization process grammatical markers may step over from a represen-
tational to an interpersonal function. For this step to occur, it is not necessary for
a marker to rst complete the entire representational chain given in (30); cross-
over may occur at any point in this hierarchy. The combination of the representa-
tional function of Deduction with the interpersonal function of Reportativity falls
therefore within the scope of our prediction.
Apart from conrming our prediction, Table 8 also shows that it is not un-
common for all four dierent evidential subcategories that we propose to occur
side by side within the same language. This is the case in 12 sample languages
from 8 dierent major families. An example of a language with all four subcatego-
ries is Desano:
() Desano
Bãdu yɨ tĩg ɨ-re paa- Reportativity
Manuel 1. brother- hit-.3..
‘Manuel hit my older brother.’ (hearsay)
(Miller 1999: 66)
() Desano
 ʔ ĩ yoaro-geaʔhra-y-a Inference
2. far- come--3
‘You must have come a long way.’ (based on what I know of your habits.)
(Miller 1999: 67)
() Desano
pisadãwai-re ba-di-g ɨ árĩ- Deduction
cat sh- eat--. be-.3..
‘The cat must have eaten the sh.’ (you can see his paw marks on the ground
where he ate it).
(Miller 1999: 68)
() Desano
g ɨa õ-ge-re era- Event Perception
1.. here-- arrive-3..
‘We arrived here.’
(Miller 1999: 65)
The data presented in this section lend strong support for the evidentiality hier-
archy in (30). Other such hierarchies have been presented in the literature and we
will briey compare them with ours here.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
Willett (1988) presents the hierarchy in (48):
() attested  reported  inferring
In this hierarchy the category reported is in the wrong position. As Table 7 shows,
there are languages which have event perception (Willett’s attested subcategory),
deduction, and possibly inference (the latter two falling within Willett’s inferring
subcategory), but no reportedness (Willett’s reported category). Cases in point
are Apalaí and Matses.
de Haan (1998) presents the hierarchy in (49):
() visual  non-visual  inferential  quotative
The rst three subcategories mentioned here seem to have a position compatible
with our hierarchy, as de Haan’s visual and non-visual subcategories fall within
our subcategory of event perception, while de Haan’s inferential subcategory
covers our deduction and inference subcategories. His quotative subcategory,
however, which corresponds to our subcategory of reportativity, is in the wrong
position if checked against our data. Our sample contains many languages with
just reportativity within their evidential system, and others that do have event
perception and reportativity but no deduction or inference.
A hierarchy proposed in the literature that comes close to ours is the one
presented in Faller (2002). It is given in (50):
()
Faller (2002) argues that there should be two implicational pathways, both from
visual to assumption, one passing through the area of reportativity, and another
through the remaining categories. The second pathway is compatible to a high
extent with our hierarchy, as Faller’s rst three subcategories (visual, audito-
ry,other sensory) fall within our subcategory of event perception, her subcate-
gory of inference on the basis of resulting evidence (inf-result) corresponds
with our subcategory of deduction, and her subcategory of inference on the
basis of reasoning (inf-reason) corresponds to our subcategory of inference.
Faller herself raises the question whether assumption properly belongs in an
evidential hierarchy or should rather be placed in a modal hierarchy. Faller’s
rstpathway runs from her subcategory visual to second hand reportative (sec-
ond) and third hand reportative (third). She tentatively poses a further possible
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
connection between reportatives and assumption. This implicational relation is
contradicted by our data. As in the case of Willett’s and de Haan’s hierarchies,
counterexamples are those in which a language has reportativity but no event-
perception in its evidential system. Nevertheless, it is interesting to see that Faller
creates a separate pathway for reportativity thus attempting to solve the diculty
of integrating it in an evidential hierarchy. We take one step further and claim it
does not belong in the hier archy at all, as it originates at a dierent level of the
grammar.
.Evidential systems
The restrictions on the co-existence of evidential subcategories also restrict the
number of possible evidential systems. On the basis of the generalizations that
can be derived from the data presented in Section 7.2 we may conclude that Table
9 provides a complete overview of possible evidential systems in terms of the four
subcategories that we distinguish, ignoring possible further subdistinctions
within each category. Systems are presented in pairs in Table 9, such that in each
case variants of systems with and without reportativity are juxtaposed. For the
sake of completeness we also include the system characterized by the absence of
evidentiality markers.
Aikhenvald (2004) is the rst publication in which a systematic classication
of evidential systems is attempted. It is therefore interesting to compare our clas-
sication with hers. The result of this comparison is given in Table 10. Aikhenvald
Table 9: Classication of evidential systems and their manifestation in the sample
Level Representational Interpersonal # lgs in
sample
Evidential
system
Event perception Deduction Inference Reportativity
a + + + + 
b + + +
a + + +
b + +
a + +
b +
a – +
b – 
Total 
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
(2004) uses the number of evidential choices in a language as the point of depar-
ture, A corresponding to a language with two evidential choices, B to one with
three choices, C to one with four and D to one with ve. Since our classication is
based on the subcategories represented and not on the number of actual choices
within each of these, various categories from Aikhenvald’s classication may
correspond to one of ours. For instance, corresponding to our evidential system
2a we nd Aikhenvald’s systems B1 (direct, inferred, reported), B4 (non- visual,
inferred, reported), C1 (visual, non-visual, inferred, reported), and C3 (direct, in-
ferred, reported, quotative). Note that the terms used here are the ones Aikhen-
vald applies as explained in Table 4. To give just one example of how these relate
to our subcategories: in Aikhenvald’s system C1 visual and non-visual are in-
stances of event perception, inferred corresponds to our deduction, and reported
to our reportativity. Overall, all Aikkenvald’s systems t into our more gener-
alizedclassication.
It is interesting to note that there are no instances of our type 1b in Aikhen-
vald’s sample, which has to do with the overall predominance of reporta-
tivity. Yet there are two languages in our sample that, on the basis of the
existingdocumentation, do not seem to have reportativity as an evidential cate-
gory: Matses and Apalaí. These two languages do not form part of Aikhenvald’s
sample.
Given the virtual omnipresence of reportativity in evidential systems, our
classication of evidential systems or the absence thereof could be reduced to the
one presented in Table 11.
We will use this classication in the following section, which explores the
areal and genealogical distribution of these systems.
Table 10: Comparison of our classication of evidential systems with that of Aikhenvald (2004)
Level Representational Interpersonal Aikhenvald
()
Evidential
system
Event perception Deduction Inference Reportativity
a + + + + C, D
b + + +
a + + + B, B, C, C
b + + B
a + + A, B
b + A, A, A
a + A, B
b –
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
. Areal and genealogical distribution
On the basis of the generalized classication of evidential systems in Table 11 the
following preliminary observations can be made with respect to the distribution
of these systems across the native languages of Brazil.
A clear genealogical pattern can be distinguished in the case of three fami-
lies. All Nambiquaran languages in the sample (Lakondê, Mamaindê, Sabanê,
Nambikuára) exhibit either system 1 or system 2. The same is true of the Tucanoan
languages (Carapana, Cubeo, Desano, Tuyuca, Wanano, Yepâ-masa) and the
Yanomam languages (Sanuma, Yanomam) in the sample.
In the case of other language families there is a mixed pattern. For instance,
in Panoan we nd systems 1 (Matses), 2 (Huariapano, Yaminahua), and 4 (Ama-
huaca, Katukina, Kaxinawá, and Shanenawa); in Tupi-Guarani systems 1 (Kamai-
urá), 3 (Guajá), and 4 (Guajajara, Guarani-Mbyá, Kokama-Kokamilla, Nheengatú,
Parintintín, Urubu-Kaapor); in Macro-Gê systems 1 (Parkatêjê) and 4 (all other
Macro-Gê languages in the sample); in Maku systems 1 (Hup), 2 (Yuhup), and 4
(Dâw, Nadëb); in Arawakan systems 1 (Tariana) and 4 (all other Arawakan lan-
guages in the sample).
It seems reasonable in these cases to look for an areal explanation of the dif-
ferent patterns within a single family. Such an explanation oers itself, for in-
stance, for the Maku family. The languages Hup and Yuhup are spoken in the vi-
cinity of Tucanoan languages, while Dâw and Nadëb are not. Indeed Epps (2008:
30) and Ospina Bozzi (2002: 65) comment on a high degree of language contact
between Hup and Yuhup respectively on the one hand, and Tucanoan languages
on the other. Vicinity to Tucanoan languages probably also explains the excep-
tionally rich evidential system of Tariana as compared to the other Arawakan
languages in the sample (Aikhenvald 2003: 7). Similar explanations are not
Table 11: Generalized classication of evidential systems and their manifestation in the sample
Level Representational Interpersonal # lgs in
sample
Evidential
system
Event perception Deduction Inference Reportativity
+ + + (+) 
+ + – (+)
+ – (+)
– (+) 
Total 
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
immediately evident in other cases, or seem to be rather unlikely, as for instance
in the case of Kamaiurá (Seki 2000: 38–39).
 Implicational relations between evidential
meanings (quantitative)
A third prediction we formulated in Section 4 is that the hierarchy in (30) will
have a quantitative correlate, such that the categories at the lower end of the hi-
erarchy will have more subdistinctions within them than the ones at the higher
end. This prediction follows from the expectation that new distinctions typically
arise at the lowest evidential layer, are closer to their lexical origin, and will only
in some cases make it to higher layers. Table 12 contains the data necessary to
check this prediction. The data underlying Table 12 are given in Table 6.
The actual numbers involved do not allow rm conclusions here, but the data
as far as available do conrm the general prediction. Along the hierarchy in (30)
the number of languages with subdistinctions decreases, and reportativity be-
haves independently of this.
The common subdivision within the subcategory of event-perception estab-
lishes an opposition between visual and non-visual perception, as illustrated
inthe Mamaindê Examples (26) and (27) in Section 3.8. This is also the case of
Wanano, Tuyuca, Kamaiurá, Tariana, Ye’pâ-masa, Lakondê and Jarawara. In all
these languages, the non-visual concerns evidence perceived by all the senses
excluding vision. An interesting dierent division occurs in Nambikuara. In this
language, it is possible to express whether an event was perceived by the speaker
alone or by the speaker and the hearer:
() Nambikuara
wa3kon3-ø-na 2hɵ̃ 3-la2
work-3.-..1-
‘He worked yesterday.’ (I tell you what I have seen.)
(Kroeker 2003: 87)
Table 12: Semantic distinctions within the subcategories of evidentiality
Level Representational Interpersonal
Evidential Event Perception Deduction Inference Reportativity
Languages with subtypes   
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
() Nambikuara
wa3ko3n-a1-tai1ti2tu3-wa2
work-1.-..1/2-
‘I worked yesterday.’ (You and I saw it)
(Kroeker 2003: 88)
The distinction between individual (53) and collective (54) also applies to Deduc-
tion in Nambikuara:
() Nambikwara
wa3kon3-ø-2hɵ̃3-la2
work-3.-.1-
‘He must have worked yesterday.’ (The deduction of the event is based on
something that the speaker saw.)
(Kroeker 2003: 87)
() Nambikwara
wa3kon3-ø-te3nait1ti2tu3-wa2
work-3.-.1/2-
‘He must have worked yesterday.’ (The event is deduced by the speaker and
the hearer based on evidence available to both.)
(Kroeker 2003: 89)
The reportative has interesting subtypes too. Apart from the Mamaindê Examples
(28) and (29) discussed above, reportative evidentiality is also split into two sub-
types in Jamamadí, which has a similar system as Mamaindê. In Nambikwara the
split is determined by the question whether the communicated content was re-
ported to just the speaker or to the speaker as well as the hearer, and in Lakondê
and Wanano the decisive factor is whether the communicated content was re-
ported by an identiable or non-identiable source.
Conclusion
In this paper we presented a new classication of evidential subcategories based
on the treatment of grammatical categories in Functional Discourse Grammar.
The resulting classication draws a sharp line between reportativity on the
one hand, and event perception, deduction, and inference on the other. The
latterthree subcategories enter into an implicational hierarchy, while reporta-
tivity forms a subcategory in its own right. We found conrmation for our
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
classicationand hierarchy in the co-existence and co-occurrence of evidential
subcategories in the languages of a broad sample of native languages of Brazil.
We furthermore showed that our hierarchy makes better predictions than exist-
ingones, mainly as a result of the separation of reportativity from all other sub-
categories of evidentiality.
Acknowledgments: We are grateful to Johan van de Auwera, Ana María Ospina
Bozzi, and four anonymous referees for their comments on earlier versions of this
paper. Marize Hattnher gratefully acknowledges nancial support from Fapesp
(grant 2009/09297-0) and CNPq (grant 314334/2009-2).
References
Aikhenvald, Alexandra Y. 2004. Evidentiality. Oxford: Oxford University Press.
Boland, Annerieke. 2006. Aspect, tense and modality: Theory, typology, acquisition. Utrecht:
LOT.
Boye, Kasper. 2010. Evidence for what? Evidentiality and scope. STUF 63. 290–307.
Bybee, Joan L. 1985. Morphology: A study of the relation between meaning and form.
Amsterdam & Philadelphia: John Benjamins.
Cinque, Guglielmo & Luigi Rizzi. 2011. The cartography of syntactic structures. In Heiko Narrog
& Bernd Heine (eds.), The Oxford handbook of linguistic analysis, 51–65. Oxford: Oxford
University Press.
Faller, Martina. 2002. Remarks on evidential hierarchies. In David I. Beaver, Luis D. Casillas
Martínez, Brady Z. Clark & Stefan Kaufmann (eds.), The construction of meaning, 37–59.
Stanford, CA: CSLI Publications.
Foley, William A. & Robert D. Van Valin Jr. 1984. Functional syntax and universal grammar.
Cambridge: Cambridge University Press.
Geurts, Bart & Janneke Huitink. 2006. Modal concord. In Paul Dekker & Hedde Zeijlstra (eds.),
Concord phenomena and the syntax semantics interface. Malaga: ESSLLI.
de Haan, Ferdinand. 1998. The category of evidentiality. Unpublished ms., University of New
Mexico.
Hattnher, Marize Mattos Dall’Aglio & Kees Hengeveld. (forthc.). The grammaticalization of
modal verbs in Brazilian Portuguese: A synchronic approach. Journal of Portuguese
Linguistics.
Hengeveld, Kees. 1989. Layers and operators in Functional Grammar. Journal of Linguistics
25(1). 127–157.
Hengeveld, Kees. 2011. The grammaticalization of tense and aspect. In Heiko Narrog & Bernd
Heine (eds.), The Oxford handbook of grammaticalization (Oxford Handbooks in
Linguistics), 580–594. Oxford: Oxford University Press.
Hengeveld, Kees, Eli Nazareth Bechara, Roberto Gomes Camacho, Alessandra Regina
Guerra, Taísa Peres de Oliveira, Eduardo Penhavel, Erotilde Goreti Pezatti, Liliane
Santana, Edson Rosa Francisco de Souza & Maria Luiza de Sousa Teixeira. 2007.
Basicillocutions in the native languages of Brazil. Alfa – Revista de Lingüística 51.2.
73–90.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
Hengeveld, Kees, Maria Luiza Braga, Elisiene de Melo Barbosa, Jaqueline Silveira Coriolano,
Juliana Jezuino da Costa, Mariana de Souza Martins, Diego Leite de Oliveira, Vinicius
Maciel de Oliveira, Luana Gomes Pereira, Liliane Santana, Cassiano Luiz do Carmo Santos
& Viviane dos Ramos Soares. 2012. Semantic categories in the indigenous languages
ofBrazil. Functions of language 19(1). 38–57.
Hengeveld, Kees & J. Lachlan Mackenzie. 2008. Functional Discourse Grammar: A typologically-
based theory of language structure. Oxford: Oxford University Press.
Lewis, M. Paul (ed.). 2009. Ethnologue: Languages of the world. 16th edn. Dallas, TX: SIL
International.
Matthews, George Hubert. 1965. Hidatsa syntax. The Hague: Mouton.
Narrog, Heiko. 2009. Modality in Japanese: The layered structure of the clause and hierarchies
of functional categories. Amsterdam & Philadelphia: John Benjamins.
Plungian, Vladimir. 2010. Types of verbal evidentiality marking: an overview. In Gabriele
Diewald & Elena Smirnova (ed.). Linguistic realization of evidentiality in European
languages, 15–58. Berlin & New York: De Gruyter Mouton.
Pollock, Jean-Yves. 1989. Verb movement, Universal Grammar and the structure of IP. Linguistic
inquiry 20. 365–424.
Roberts, Ian & Anna Roussou. 2003. Syntactic change: A minimalist approach to
grammaticalization. Cambridge: Cambridge University Press.
San Roque, Lila & Robyn Loughnane. 2012. The New Guinea Highlands evidentiality area.
Linguistic Typology 16(1). 111–167.
Torres Sánchez, Nadiezdha. 2013. La evidencialidad en las lenguas indígenas americanas:
Unenfoque areotipológico [Evidentiality in indigenous American languages: An areal-
typological study]. Mexico: Instituto Nacional de Antropología e Historia.
Willett, Thomas. 1988. A cross-linguistic survey of the grammaticalization of evidentiality.
Studies in Language 12. 51–97.
Appendix A: Abbreviations
1=rst person, 1/2=rst+second person, 2=second person, 3=third person,
=aected, =certainty, =completive, =declarative, =
deduction, =dependent, =directional, =dualis, =dynamic,
=existential, =exclusive, =feminine, =nal nominal sux, =
immediate, =imperfective, =indicative, =inferential, =
ingressive, =locative, =masculine, =negative, =non-feminine,
=nominalizer, 1=non-rst person, 3=non-third person, =
non-visual, =non-witnessed, =object, =event perception,
=perfective, =plural, =possessive,  =present, =past, =
recent, =remote, =reportative, =subject, =singular, =
specic, =visual, =verbal sux, =witnessed.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
Appendix B: Descriptions of sample languages
Amahuaca
Russel, Robert L. 1965. A transformational grammar of Amahuaca (Pano). Columbus, OH: Ohio
State University MA thesis.
Apalaí
Koehn, Edward H. & Sally S. Koehn. 1986. Apalaí. In Desmond C. Derbyshire & Georey K.
Pullum (eds.), Handbook of Amazonian languages, vol. 1, 33–127. Berlin & New York:
Mouton de Gruyter.
Apinayé
Callow, John C. 1962. The Apinayé language: Phonology and grammar. London: School
ofOriental and African Studies dissertation.
Ham, Patrícia. 1961. Apinayé grammar. Brasília: Summer Institute of Linguistics.
Ham, Patrícia & Helen Waller & Linda Koopman. 1979. Aspectos da língua Apinayé [Aspects
ofthe Apinayé language]. Cuiabá, MT: Sociedade Internacional de Lingüística.
Koopman, Linda. 1976. Cláusulas semânticas na língua Apinajé [Semantic clauses in the
Apinajé language]. Série lingüística 5. 301–330.
Aweti
Borella, Cristina Cássia. 2000. Aspectos morfossintáticos da língua Aweti [Morphosyntactic
aspects of the Aweti language]. Campinas: Universidade Estadual de Campinas MA thesis.
Baré
Aikhenvald, Alexandra Y. 1995. Bare. Munich: Lincom Europa.
Carapana
Metzger, Ronald G. 1981. Gramática popular del Carapana. Bogotá: Ministerio de Gobierno.
Cubeo
Morse, Nancy L. & Michael B. Maxwell. 1999. Cubeo grammar. Arlington, TX: Summer Institute
of Linguistics and the University of Texas at Arlington.
Dâw
Martins, Silvana Andrade. 2004. Fonologia e gramática Dâw [Phonology and grammar of Dâw].
Amsterdam: Vrije Universiteit Amsterdam dissertation.
Desano
Miller, Marion. 1999. Desano grammar. Arlington, TX: Summer Institute of Linguistics and the
University of Texas at Arlington.
Fulniô
Meland, Douglas. 1968. Fulniô grammar. Brasilia: Summer Institute of Linguistics.
Guajá
Magalhães, Marina Maria Silva. 2007. Sobre a morfologia e a sintaxe da lingua Guajá (Família
Tupí-Guaraní) [About the morphology and syntax of the Guajá language (Tupi-Guaraní
family)]. Brasilia: Universidade de Brasília dissertation.
Guajajara
Bendor-Samuel, David. 1972. Hierarchical structures in Guajajara. Norman, OK: Summer
Institute of Linguistics of the University of Oklahoma.
Guarani-Mbyá
Martins, Marci Fileti. 2003. Descrição e análise de aspectos da gramática do Guarani Mbyá
[Description and analysis of aspects of the grammar of Mbyá Guarani]. Campinas:
Universidade Estadual de Campinas dissertation.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
Guató
Palacio, Adair Pimentel. 1984. Guató: a lingua dos índios canoeiros do rio Paraguai [Guató:
Thelanguage of the canoe indians of the Paraguai river]. Campinas: Universidade Estadual
de Campinas dissertation.
Huariapano
Gomes, Graziela de J. 2010. Aspectos morfossintáticos da língua Huariapano (Pano)
[Morphosyntactic aspects of the Huariapano language (Pano)]. Campinas: Universidade
Estadual de Campinas MA thesis.
Hup
Epps, Patience. 2008. A grammar of Hup. Berlin & New York: Mouton de Gruyter.
Jamamadí
Campbell, Robert. 1977. Marcadores de fonte de informação na língua Jamamadí [Markers
ofinformation source in the Jamamadí langauge]. Série lingüística 7. 117–126.
Jarawara
Dixon, Robert M. W. 2004. The Jarawara language of Southern Amazonia. Oxford: Oxford
University Press.
Kadiwéu
Griths, Glyn & Cynthia Griths. 2006. Aspectos da língua Kadiwéu [Aspects of the Kadiwéu
language]. Cuiabá: Sociedade Internacional de Linguística.
Kamaiurá
Seki, Lucy. 2000. Gramática do Kamaiurá [A grammar of Kamaiurá]. Campinas: Editora da
Unicamp.
Kanoê
Bacelar, Laércio Nora. 2004. Gramática da língua Kanoê [A grammar of the Kanoê language].
Nijmegen: Katholieke Universiteit Nijmegen.
Karitiana
Everett, Caleb. 2006. Patterns in Karitiana: Articulation, perception and grammar. Houston,
TX:Rice University dissertation.
Karo
Gabas Jr., Nilson. 1999. A Grammar of Karo, Tupi, Brazil. Santa Barbara, CA: University of
California Santa Barbara dissertation.
Katukina (Panoan)
Aguiar, Maria Sueli. 1994. Análise descritiva da língua Katukina-Pano [A descriptive analysis of
the Katukina-Pano language]. Campinas: Universidade Estadual de Campinas dissertation.
Katukina-Kanamari
dos Anjos, Zoraide. 2011. Fonologia e Gramática Katukina-Kanamari [Phonology and grammar
of Katukina-Kanamari]. Utrecht: LOT.
Kaxinawá
Montag, Susan. 2004. Lições para a aprendizagem da língua Kaxinawá [Materials for the study
of the Kaxinawá language]. Lima: Ministerio de Educación.
Kokama-Kokamilla
Yopán, Rosa V. 2010. A Grammar of Kokama-Kokamilla. Eugene, OR: University of Oregon
dissertation.
Kuikuro
Franchetto, Bruna. 2002. Kuikuro: uma língua ergativa no ramo meridional da Família Karib
(Alto Xingu) [Kuikuro: An ergative language of the southern branch of the Carib family
(AltoXingu)]. In Francisco Queixalós (ed.), Ergatividade na Amazônia I, 15–44.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
Paris:Centre d’études dês languages indigènes d’Amerique (CNRS,IRD) and Brasilia:
Laboratório de Línguas Indígenas,.
Santos, Gelsama M. F. 2007. Morfologia Kuikuro: gerando nomes e verbos [The morphology
ofKuikuro: Generating nouns and verbs]. Rio de Janeiro: Universidade Federal do Rio
deJaneiro dissertation.
Kwaza
van der Voort, Hein. 2004. A Grammar of Kwaza. Berlin & New York: Mouton de Gruyter.
Lakondê
Telles, Stella & Leo Wetzels. 2006. Evidentiality and epistemic mood in Lakondê. In Grażyna
J.Rowicka & Eithne B. Carlin (eds.), What’s in a verb? Studies in the verbal morphology
ofthe languages of the Americas, 235–252. Utrecht: LOT.
Macushi
Abbott, Miriam. 1991. Macushi. In Desmond C. Derbyshire & Georey K. Pullum (eds.),
Handbook of Amazonian languages, vol. 3, 23–160. Berlin: Mouton de Gruyter.
Mamaindê
Eberhard, David Marker. 2009. Mamaindê Grammar: A Northern Nambikwara language and its
cultural context. 2 vols. Utrecht: LOT.
Matses
Fleck, David. 2003. A grammar of Matses. Houston, TX: Rice University dissertation.
Maxacalí
Pereira, Deuscreide Gonçalves. 1992. Alguns aspectos gramaticais da língua Maxacalí [Some
grammatical aspects of the Maxacalí language]. Belo Horizonte: Universidade Federal
deMinas Gerais MA thesis.
Araújo, Gabriel Antunes. 2000. Fonologia e morfologia da língua Maxacalí [Phonology and
morphology of the Maxacalí language]. Campinas: Universidade Estadual de Campinas
MAthesis.
Moré
Ferrarezi, Junior Celso. 1997. Nas águas dos Itenês: um estudo semântico com a língua Moré
[Atthe Itenês waters: A semantic study of the Moré language]. Campinas: Universidade
Estadual de Campinas MA thesis.
Munduruku
Gomes, Dioney Moreira. 2006. Estudo morfológico e sintático da língua Mundurukú (Tupi)
[Amorphological and syntactic study of the Mundurukú language]. Brasilia: Universidade
de Brasília dissertation.
Nadëb
Weir, E. M. H. 1986. A negação e outros tópicos da gramática Nadëb [Negation and other
topics in the grammar of Nadëb]. Campinas: Universidade Estadual de Campinas MA
thesis.
Nambikuára
Kroeker, Menno H. 2003. Gramática descritiva da língua Nambikuára [A descriptive grammar
ofthe Nambikuára language]. Cuiabá: Sociedade Internacional de Linguística.
Nheengatú
Cruz, Aline da. 2011. Fonologia e gramática do Nheengatú [Phonology and grammar
ofNheengatú]. Utrecht: LOT.
Palikur
Aikhenvald, Alexandra Y. & Diana Green. 1998. Palikur and the typology of classiers.
Anthropological Linguistics 40(3): 429–480.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Evidentiality in the native languages of Brazil
Dooley, Robert A. & Harold G. Green. 1977. Aspectos verbais e categorias discursivas da língua
Palikur [Verbal aspects and discourse categories in the Palikur language]. Série lingüística
7. 7–28.
Parecis
Rowan, Orland & Eunice Burgess. 1979. Gramática Parecis [Parecis grammar]. Brasilia: Summer
Institute of Linguistics.
Parintintín
Pease, Helen. 1968. Parintintín grammar. Brasilia: Summer Institute of Linguistics.
Parkatêjê
Ferreira, Marília de Nazareth de Oliveira. 2003. Estudo morfossintático da língua Parkatêjê
[Amorphosyntactic study of Parkatêjê]. Campinas: Universidade Estadual de Campinas
dissertation.
Pirahã
Everett, Daniel. 1986. Pirahã. In Desmond C. Derbyshire & Georey K. Pullum (eds.), Handbook
of Amazonian languages, vol. 1, 200–325. Berlin & New York: Mouton de Gruyter.
Pykobje
Amado, Rosane de Sá. 2004. Aspectos morfofonológicos do Gavião-Pykobjê
[Morphophonological aspects of Gavião-Pykobjê]. São Paulo: Universidade de São Paulo
dissertation.
Rikbáktsa
Silva, Leia de J. 2005. Aspectos da fonologia e da morfologia da língua Rikbáktsa [Aspects
ofthe phonology and morphology of the Rikbáktsa language]. Brasilia: Universidade de
Brasília dissertation.
Sabanê
Araújo, Gabriel A. 2004. A grammar of Sabanê, a Naimbikwaran language. Amsterdam: LOT.
Sanuma
Borgman, Donald M. 1990. Sanuma. In Desmond C. Derbyshire & Georey K. Pullum
(eds.),Handbook of Amazonian languages, vol. 2, 15–248. Berlin & New York:
Mouton de Gruyter.
Shanenawa
Cândido, Gláucia Vieira. 2004. Descrição morfossintática da lingua Shanenawa (Pano)
[Amorphosyntactic description of the Shanenawa language (Pano)]. Campinas:
Universidade Estadual de Campinas dissertation.
Surui
Bontkes, Willem & Robert A. Dooley. 1985. Verication particles in Surui. Brasilia: Summer
Institute of Linguistics
Tariana
Aikhenvald, Alexandra Y. 2003. A Grammar of Tariana, from northwest Amazonia. Cambridge:
Cambridge University Press.
Terena
Ekdahl, Elizabeth M. & Nancy E. Butler. 1979. Aprenda Terena [Learn Terena]. Brasilia: Summer
Institute of Linguistics.
Tiriyó
Meira, Sérgio. 1999. A grammar of Tiriyó. Houston, TX: Rice University dissertation.
Trumaí
Guirardello, Raquel. 1999. A reference grammar of Trumaí. Houston, TX: Rice University
dissertation.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
Kees Hengeveld and Marize Mattos Dall’Aglio Hattnher
Tuyuca
Barnes, Janet. 1984. Evidentials in the Tuyuca verb. International Journal of American
Linguistics 50. 255–271.
Urubu-Kaapor
Kakumasu, James Y. 1986. Urubu-Kaapor. In Desmond C. Derbyshire & Georey K. Pullum (eds.),
Handbook of Amazonian languages, vol. 1, 326–403. Berlin & New York: Mouton deGruyter.
Waiwai
Hawkins, Robert E. 1998. Wai Wai. In Desmond C. Derbyshire & Georey K. Pullum (eds.),
Handbook of Amazonian languages, vol. 4, 25–224. Berlin & New York: Mouton deGruyter.
Wanano
Stenzel, Kristine S. 2004. A reference grammar of Wanano. Boyulder, CO: University of Colorado
dissertation.
Wapixana
dos Santos, Manoel G. 1992. Uma gramática do Wapixana (Aruák): Aspectos da fonologia,
damorfologia e da sintaxe [A grammar of Wapixana (Arawak): Aspects of phonology,
morphology and syntax]. Campinas: Universidade Estadual de Campinas dissertation.
Xavante
de Oliveira, Rosana Costa. 2007. Morfologia e sintaxe da língua Xavante [Morphology and
syntax of the Xavante language]. Rio de Janeiro: Universidade Federal do Rio de Janeiro
dissertation.
Mcleod, Ruth & Valerie Mitchel. 2003. Aspectos da língua Xavante [Aspects of the Xavante
language]. Cuiabá: Sociedade Internacional de Lingüística.
Yaminahua
Faust, Norma & Eugene E. Loos. 2002. Gramática del idioma Yaminahua. Lima: Instituto
Lingüístico de Verano.
Yanomamɨ
Ramirez, Henri. 1994. Le parler Yanomamɨ des Xamatauteri. Aix-en-Provence: Université
deProvence dissertation.
Ye’pâ-masa
Ramirez, Henri. 1997. A fala Tukano dos Ye’pâ-Masa. Gramática, dicionário, método de
aprendizagem [The Tucano language of the Ye’pâ-Masa. Grammar, dictionary, language
course]. 3 vols. Manaus: Inspetoria Salesiana Missionária da Amazônia.
Yuhup
Ospina Bozzi, Ana Maria. 2002. Les structures élémentaires du Yuhup Maku: langue de
l’Amazonie colombienne: Morphologie et syntaxe. Paris: Université Paris 7 dissertation.
Authenticated | p.c.hengeveld@uva.nl author's copy
Download Date | 4/30/15 5:29 PM
... Afin d'analyser des formes polysémiques qui expriment ces deux sous-types en portugais, cette étude travaille avec des occurrences des verbes de perception ver, observar, ouvir, sentir et perceber, , retirées de blogs de thèmes variés, en qui ces verbes fonctionnent comme la tête d'une proposition simple ou complexe et en qui l'élément sous la portée du médiatif fonctionne comme un complément de cette proposition. Premièrement, les résultats corroborent les affirmations de travaux antérieurs (HATTNHER, 2013;HATTNHER, 2015) selon lesquelles la perception d'événement est sémantiquement caractérisée par la simultanéité de la perception et de l'événement perçu, alors que la déduction est caractérisée par un éloignement temporel et/ou spatial entre la déduction et l'événement déduit. Deuxièmement, dans les données analysées, avec des compléments non propositionnels qui codifient des unités sémantiques de tête lexicale, il n'est trouvé que des propositions qui véhiculent la perception d'événement. ...
Article
Full-text available
p>A evidencialidade é uma categoria qualificacional que marca a fonte da informação contida em um enunciado ou o meio pelo qual o falante obteve essa informação. Adotando o arcabouço teórico da Gramática Discursivo-Funcional (HENGEVELD; MACKENZIE, 2008) e a classificação da evidencialidade proposta por Hengeveld e Hattnher (2015) e Hengeveld e Fischer (2018), é possível reconhecer cinco subtipos evidenciais, dois dos quais interessam a este trabalho: a dedução , que é um raciocínio embasado por uma percepção sensorial, e a percepção de evento , que diz respeito à atestação de um estado de coisas por meio de um dos sentidos. No intuito de analisar formas polissêmicas que expressam ambos os subtipos evidenciais em português, este estudo trabalha com ocorrências dos verbos de percepção ver , observar , ouvir , sentir e perceber , retiradas de blogs de temáticas diversificadas, em que esses verbos funcionam como núcleo de uma oração simples ou complexa e em que o elemento sob o escopo do evidencial funciona como complemento dessa oração. Em primeiro lugar, os resultados corroboram afirmações de trabalhos anteriores (HATTNHER, 2013; HENGEVELD; HATTNHER, 2015) de que a percepção de evento é caracterizada, semanticamente, pela simultaneidade entre a percepção e o evento percebido, ao passo que a dedução é caracterizada por um distanciamento temporal e/ou espacial entre a dedução e o evento deduzido. [...] Por fim, o evidencial de percepção de evento é encontrado apenas no passado e no presente, assim como o evento sob o escopo do evidencial dedutivo. PALAVRAS-CHAVE : Evidencialidade. Dedução. Percepção de evento. Gramática Discursivo-Funcional.</p
Article
Full-text available
This article addresses the question of how the distribution and role of English evidential - ly adverbs in the noun phrase can be accounted for using the framework of Functional Discourse Grammar (FDG). Both adverbs and adjectives occurring in noun phrases are categorized in various ways. The results of the categorization offer insights into the distribution of these adverbs and adjectives. Four generalizations are arrived at concerning the combination of evidential adverbs and adjectives in noun phrases. First, the lower in the FDG hierarchy the category of an adverb, the less frequent the occurrence of that category in the noun phrase. Thus, higher reportative adverbs are very frequent, and lower adverbs of event perception are very infrequent. Second, evidential adverbs do not modify adjectives that express the speaker’s subjective evaluation of the referent. Third, the higher-level reportative and inferential adverbs modify adjectives expressing permanent properties, whereas the lower adverbs of deduction and event perception do not. Finally, neither restrictiveness nor the evaluative vs descriptive nature of the adjective appears to solely determine the category of evidential modification of the adjective. We furthermore discuss the pragmatic effects of the evidential adverb in the noun phrase, such as distancing, and the stress shift that may accompany it.
Article
Full-text available
O obxectivo deste artigo é presentar unha visión xeral do sistema evidencial galego, nomeadamente das fontes indirectas inferencial e reportativa. Tras levar a cabo unha revisión dos estudos previos dos adverbios disque e seica, analizamos os dous significados evidenciais mediante exemplos do conxunto de marcadores que propoñemos e amosamos casos nos que a fonte é indistinta entre os dous valores canónicos previos, o que se vén en chamar “significado indirecto-indiferente”. Comprobamos que as expresións evidenciais actúan como unidades pragmáticas polifuncionais transmisoras de diversos matices sociais relacionados coas estratexias de cortesía negativa (atenuación), autoridade epistémica e establecemento dun territorio de información común compartido entre falante e oínte.
Article
Full-text available
Analisando orações complexas em que a matriz é nucleada pelos verbos inferir, deduzir, concluir, perceber, observar e ver, este trabalho objetiva testar a hipótese de que, se esses verbos expressam inferência ou dedução, subtipos evidenciais de acordo com Hengeveld e Hattnher (2015), o tempo e o modo da encaixada é diferente a depender do subtipo. Os resultados comprovam essa hipótese, já que, nos dados analisados, a encaixada com a informação deduzida só se apresenta em tempos do passado e do presente do indicativo, enquanto a informação inferida ocorre nos tempos presente, passado e futuro e nos modos indicativo, subjuntivo e condicional. As ocorrências provêm do Corpus Brasileiro (SARDINHA; MOREIRA FILHO; ALAMBERT, 2010) e do Timestamped JSI web corpus 2014-2020 (TRAMPUS; NOVAK, 2012), e são analisadas com o auxílio do Goldvarb X (SANKOFF; TAGLIAMONTE; SMITH, 2005).
Article
While evidential categories have similar meanings and uses cross-linguistically, the names given to the categories tend to vary. This article surveys the terms for evidentials used in descriptive grammars of 51 languages that have at least three evidential categories. Comparing terms to find common terminological tendencies can provide useful insights into the descriptions of categories. There is great deal of term synonymy and term polysemy, but each evidential category has its typical terms that are semantically motivated. The term synonymy is the result of emphasizing different semantic aspects of the evidential categories, which leads to variable term choices. There is terminological overlap between categories involving sensory perception (i.e. visual, direct, and nonvisual sensory evidentials), inference, and second-hand information, which suggests that they may not always be considered sufficiently conceptually distinct to merit unique terms, especially when the terms for inferential and assumptive evidentials are frequently created from the same words. The variation of the terminology is also connected to how the understanding of evidentiality and relationship to other adjacent categories has developed.
Article
While evidential categories have similar meanings and uses cross-linguistically, the names given to the categories tend to vary. This article surveys the terms for evidentials used in descriptive grammars of 51 languages that have at least three evidential categories. Comparing terms to find common terminological tendencies can provide useful insights into the descriptions of categories. There is great deal of term synonymy and term polysemy, but each evidential category has its typical terms that are semantically motivated. The term synonymy is the result of emphasizing different semantic aspects of the evidential categories, which leads to variable term choices. There is terminological overlap between categories involving sensory perception (i.e. visual, direct, and nonvisual sensory evidentials), inference, and second-hand information, which suggests that they may not always be considered sufficiently conceptually distinct to merit unique terms, especially when the terms for inferential and assumptive evidentials are frequently created from the same words. The variation of the terminology is also connected to how the understanding of evidentiality and relationship to other adjacent categories has developed.
Article
In contrast to the specialized literature (Hardman 1986, 2001; Martinez-Vera 2020), this paper questions the direct evidential function of the sentence-type suffix -wa in Northern Aymara (Cerr n-Palomino 2000), an understudied Andean language spoken in the area around Lake Titicaca. By taking a narrow perspective on evidentiality, it proposes that -wa is a modal epistemic marker that indicates the speaker’s high commitment to information. Semantic considerations that support this proposal focus on both the interplay of -wa with other grammatical categories and its distribution within the sentence. Additionally, considerations on the scope properties of the suffix, as applied by Hengeveld & Dall’Aglio-Hatthner (2015), are taken into account in the discussion.
Preprint
Full-text available
The class of adverbs contains a large number of subcategories, ranging from adverbs of manner and degree to modal and illocutionary adverbs. What all adverbs have in common is that they modify a non-nominal head. This head may be lexical in nature, such as an adjective modified by a degree adverb, or syntactic, such as a full clause modified by an evidential adverb. In this chapter, adverbs are classified in terms of the types of head they modify, which will lead to a systematic classification of adverb classes that have been identified individually in the literature. It is furthermore shown that this classification provides the basis for a series of generalizations concerning the form and behaviour of classes of adverbs. These generalizations are formulated using the theoretical framework of Functional Discourse Grammar.
Article
Full-text available
The article presents the first survey of grammaticized evidentiality in a clus-ter of languages spoken in Papua New Guinea, including the Ok-Oksapmin, Duna-Bogaia, Engan, East and West Kutubuan, and Bosavi families. We com-pare certain features of these languages and outline how they contribute to the typological understanding of evidentiality. Findings concern the underex-plored category of participatory evidentiality, the morphological form of direct versus indirect evidentials, relationships between person, information source, and time, and complex treatments of the "perceiver" role implied by eviden-tials. The systems of the area are rich and varied, providing great scope for further descriptive and typological work.
Article
Full-text available
This article investigates which semantic categories, as defined in Functional Discourse Grammar, formally manifest themselves in a sample of native languages of Brazil, and the extent to which the distribution of these manifestations across categories can be described systematically in terms of implicational hierarchies. The areas subjected to investigation are basic interrogative words, basic demonstrative words, and nominalization strategies.
Article
This paper deals with the scope properties of evidential meanings. It rejects the idea that different types of evidential meanings have different scope properties. More basically, it rejects the idea that evidential meanings apply to ‘speech acts’ or to ‘states of affairs’. The paper argues that evidential meanings share scope properties in the sense that they are all conceptually dependent on a ‘proposition’ – i.e. a meaning unit which can be said to have a truth value. Subsequently, it outlines how the scope properties can be employed in criteria of membership of the category of evidentiality.
Article
Não surpreende que determinados aspectos temporais costumem ocorrer com maior freqüência em alguns trechos de uma narrativa que em outros. Na montagem ou ambiente criado, por exemplo, costumamos deparar com aspectos que comunicam duração extensa (como no caso dos verbos progressivos em português: 'estar' -ndo), pois supõe-se que as condições descritas vigorem por muito tempo. É por isso que uma categoria informacional ao nível discursivo pode sugerir determinado aspecto sem excluir outros aspectos também viáveis. Na língua Palikur1, é altamente previsível a distribuição dos cinco aspectos verbais dentro da narrativa: os tre s que não comunicam a terminação de um evento são previsíveis na base semantica do contexto imediato, ao passo que os dois que transmitem tal terminação podem ser prognosticados na base das categorias no nível discursìvo. Por isso, os marcadores que significam aspecto temporal no nível clausal podem também assumir outra função no nível discursivo. Isso não implica, porém, em perda da realidade psicológica do seu aspecto temporal; muito pelo contrário, como se verá na seção 5, os efeitos psicologicos do aspecto temporal podem ser de fato a base das suas funções no nível discursivo. Após uma breve descrição do sistema aspectual da língua Palikur2 e uma vista panoramica de um sistema de categorias relevantes ao nível discursivo, iremos ilustrar a relação entre estes dois sistemas. Especificamente, o aspecto completivo Palikur é a forma verbal fundamental (i. e., a comumente usada) para informação discursiva primária, sendo fundamental o aspecto unitário para informação secundária. Tal relação esclarece não somente as possíveis implicações do aspecto verbal como também proporciona evidência da realidade lingüística de categorias informativas no nível discursivo (v. também Popjes, 1974).