Content uploaded by Karolina Grzech
Author content
All content in this area was uploaded by Karolina Grzech on Oct 13, 2023
Content may be subject to copyright.
Egophoricity and evidentiality: Dierent
categories, similar discourse functions
Insights on conversational data from the Tibetan
Plateau and the Amazonian Foothills
Erika Sandmanand Karolina Grzech,
University of Helsinki |University of Valencia |Stockholm
University
This article discusses how evidential and egophoric making is used to man-
age knowledge in interaction. To this end, it analyzes interactional data
from Wutun (mixed Sinitic, Northwest China) and Upper Napo Kichwa
(Quechuan, Ecuador). Wutun has an egophoric marking system, which,
according to the denition of egophoricity, encodes involvement/lack of
involvement in the described event. Upper Napo Kichwa has a set of evi-
dentials, which, according to theory, encode the source of evidence for a
given proposition. The two languages are typologically unrelated. However,
when we look closely at how speakers of Wutun and Kichwa use epistemic
markers, we discover functional similarities not predicted by the dominant
denitions of epistemicity and evidentiality. In both languages, the use of
the markers is conditioned by the interpersonal context of the interaction,
and speakers use egophoric and evidential marking to signal their epistemic
rights and responsibilities with respect to other speech-act participants.
Keywords: epistemicity, Conversation Analysis, Kichwa, Wutun, stance
. Introduction
The aim of our paper is to discuss how evidential and egophoric markers are used
to manage knowledge states and asymmetries in moment-by-moment social inter-
action. We draw on analytical concepts of epistemicity developed in Conversation
Analysis (e.g., Stivers et al. a; Heritage a; Heritage & Raymond ) and
Interactional Linguistics (Couper-Kuhlen & Selting ) and apply these con-
cepts to rst-hand eld data from two lesser-studied languages. We discuss how
to establish and analyse comparative contexts in unrelated languages, using data
https://doi.org/10.1075/il.21014.san |Published online: 22 August 2022
Interactional Linguistics 2:1 (2022), pp. 79–109. ISSN 2666-4224 |E‑ISSN 2666-4232
© John Benjamins Publishing Company
from spoken interaction. Our paper contributes to research on epistemicity by
comparing discourse functions of epistemic marking in the two less-studied lan-
guages.
Wutun (mixed Sinitic language, Northwest China) and Upper Napo Kichwa
(Quechuan, Ecuador) are two unrelated, typologically distinct languages, spoken
in dierent parts of the world. Wutun is spoken by ca. people in Qinghai
Province, Northwest China, in the context of a linguistic area known as Amdo
Sprachbund or Qinghai-Gansu Sprachbund (see Janhunen ; Slater ). In
terms of phonology, lexicon, and grammar it represents a unique combination of
Chinese and Tibetan features that are due to long-term language contact between
Northwest Mandarin and Amdo Tibetan (Sandman ). While most of the
Wutun basic vocabulary and grammatical morphemes have cognates in other
varieties of Mandarin Chinese, the morphosyntax is very un-Chinese and repli-
cates structural features of Tibetic languages. Typologically, Wutun is a head-nal
SOV-language with agglutinative, mainly suxing morphology.
Upper Napo Kichwa (henceforth UNK) is a Quechuan language spoken by
ca. people (INEC ) in the Ecuadorian Amazon. According to the
widely-used classication of Quechuan languages (Torero ) it belongs to the
Quechua IIB branch. Most Quechuan languages are spoken in the Andes, but
UNK is one of a few Quechuan varieties spoken in the Amazon. The culture
of UNK speakers resembles other Amazonian cultures from the region, but the
structure of the language is typical of the Quechuan family. UNK is agglutinative,
almost exclusively suxing, and verbal morphology is the most complex part of
its morphosyntax.
Both Wutun and UNK exhibit epistemic marking systems, and we look at the
two languages jointly because of functional similarities between these systems.
Wutun marks egophoricity (i.e., ‘personal involvement or privileged access of a
speaker’, San Roque et al. :), while UNK features markers that are cognates
of evidentials (i.e., marking of ‘the source of information’, e.g., Aikhenvald )
in other Quechuan varieties. However, interactional analysis of the two languages
points to marked similarities between how these two categories are used; both
are connected to the expression of epistemic status and stance (Heritage a;
b), as well the assignment of epistemic authority (Heritage & Raymond ;
Stivers et al. b).
Most studies on epistemicity in social interaction are based on European lan-
guages, which lack highly grammaticalized epistemic systems and mainly express
evidential-like functions with evidential verbs, adverbs, or reported speech con-
structions (cf. Floyd ). While a number of studies on evidentiality in non-
European languages look at evidentials in their discourse contexts and consider
the pragmatics of evidentials (Mushin ; San Roque ; Schultze-Berndt
Erika Sandman and Karolina Grzech
; Michael , ; Nuckolls & Michael amongst others), studies
on under-documented languages with highly grammaticalized evidential systems
are rarely conducted by using methods informed by ndings from Conversation
Analysis (but see Gipper ; ; ; Grzech ) or Interactional Linguis-
tics.
Our article highlights the relevance of stance to the use of evidential expres-
sions in interaction (see Section ). Non-grammaticalized and grammaticalized
evidentials have been studied within dierent research traditions, and this might
be the reason why some authors working on evidentiality claim that stance would
be more important for the use of non-grammaticalized evidentials. For example,
Aikhenvald () analyses grammaticalized evidentials in terms of information
source and clearly distinguishes them from non-grammaticalized ‘evidential
strategies’. This approach has inuenced many descriptions of evidential systems.
For instance, Cli () has shown that reported speech in English conversation
has an important role in indexing stance in specic sequential contexts, as in case
of competing assessments in which speakers use reported speech to make their
own assessments more powerful. Cli labels this non-grammaticalized evidential
expressions with a stance-taking function as ‘interactional evidentials’ to distin-
guish it from grammaticalized ‘stand-alone evidentials’, which indicate informa-
tion source (Cli :–). However, our study provides further evidence
to the point already made by Mushin () that stance is also relevant in lan-
guages that have highly grammaticalized epistemic systems, and that – even in
those cases – the choice of an epistemic marker in spoken interaction cannot be
explained exclusively in terms of information source.
Our paper also touches on the question of the relationship between eviden-
tiality and egophoricity, which has been a matter of debate during the last decades.
In typological literature, egophoricity and evidentiality are oen treated as con-
ceptually distinct categories (Aikhenvald , ; Aikhenvald & Dixon ;
Aikhenvald & LaPolla ; Widmer ), and their established denitions point
to involvement (in case of egophoricity) and access to evidence (in case of eviden-
tiality). However, the complexity found in the descriptive data is not reected in
how egophoricity and evidentiality are dened (cf. Bergqvist & Grzech in press).
We demonstrate that taking the interactional workings of epistemic markers as a
point of comparison makes it possible to compare languages analysed with evi-
dential and egophoric marking systems in a meaningful way, and sheds new light
on the relationship between evidentiality and egophoricity. The paper focusses on
the Wutun EGO marker and UNK ‘direct evidential’, showing that they are used
to achieve similar purposes in discourse.
The structure of the paper is as follows: In Section , we review the existing
denitions of evidentiality and egophoricity, and discuss their relevance to our
Egophoricity and evidentiality
analysis. In Section we dene the relevant interactional concepts used in this
study. In Sections and we present a detailed analysis of Wutun and UNK epis-
temic marking based on data drawn from spoken interaction. Finally, we discuss
our ndings in Section and provide some concluding remarks in Section .
. Evidentiality and egophoricity
The conceptualization of evidentiality as ‘source of information’(Aikhenvald
, ; Aikhenvald & Dixon ; Willett ) indicates that evidential
marking indexes a link between the speaker’s utterance and the objective,
language-external reality. The conceptualization focusing on ‘mode of access’
(Michael ; Cornillie ) brings to the fore the speaker’s subjective concep-
tualization of evidence, which allows for the possibility that interlocutors might
access and interpret the same language-external reality in dierent ways, or that
dierent ways of accessing an event are possible even in case of the same ‘source
of information’. At the same time, the ‘source of information’ and ‘mode of access’
denitions share the assumption that evidentiality indexes a relationship between
the text-external world and the speaker, implicitly assuming other speech-act par-
ticipants are not relevant.
As more and more languages featuring dedicated evidential systems are being
described, it becomes apparent that evidential marking indexes not only the
source of evidence/mode of access, but also the distribution of knowledge
between speech-act participants, and, in many cases, knowledge-related rights
and responsibilities (cf. e.g., Nuckolls & Michael ; Grzech a). Evidentials
including intersubjective meaning components have also been described, for
example, in Kogi (Arawako-Chibchan, Colombia, Bergqvist ), Yurakaré (iso-
late, Bolivia, Gipper ; ) or, beyond South America, in Jamingjung/Ngali-
wurru (Mirndi, Australia, Schultze-Berndt ). The currently available
descriptive data indicates that, in order be empirically-grounded, the conceptual-
ization of evidentiality needs to go beyond the relationship between the speaker
and the text-external world, and to also include the intersubjective component:
how knowledge, or access to knowledge, is distributed between the speaker and
the addressee (cf. e.g., Bergqvist & Kittilä ). This links evidentiality to
egophoricity.
Egophoricity is usually dened as grammatical encoding of personal involve-
ment or privileged access of a speaker in a represented event or situation
(Bergqvist & Knuchel ; San Roque et al. :). Markers in egophoric sys-
tems interface with the speech-act roles of the speaker (rst person), the addressee
(second person), and the other (third person). In canonical egophoric marking
Erika Sandman and Karolina Grzech
systems, the same marker (EGO) is associated with rst person in statements, sec-
ond person in questions, and with the reported speaker in reported speech. The
other marker (NON-EGO) is used elsewhere. This shiability of egophoric mark-
ing is connected to the principle of privileged epistemic access, whereby the role
of the primary knower is associated with dierent speech act roles in dierent
contexts. In statements, the primary knower is usually the speaker; in questions,
the privileged epistemic access shis to an addressee. In reported speech, author-
ity is indexed to the reported speaker.
Many Tibetic languages (Garrett ; Tournadre & LaPolla ; DeLancey
) and egophoric-marking languages of Western China (Slater ; Sandman
) have highly exible systems in which egophoric marking is not tied to per-
son or sentence type and can express a wide range of interpersonal stances, such
as intentionality, certainty, assertiveness, agency, annoyance, and irony. This is
also the case in Wutun, the egophoric marking system of which is very exible
and oen deviates from canonical egophoric marking pattern, especially in every-
day conversations.
The core dening parameter relevant to egophoric systems – personal
involvement – can also be understood as a mode of access to information, and
according to this perspective, egophoricity ts within the evidential framework
(see Section ). This is recognized in many studies on evidentiality, in which
markers of personal involvement are labelled as participatory or performative evi-
dentials (e.g., Oswalt ; Mithun ; Loughnane ). Despite that, Aikhen-
vald (, :) denounces ‘participatory’ or ‘performative’ marking as
quasi-evidentials, claiming that their descriptions result from a confusion
between evidentiality and egophoricity (information source and access to infor-
mation).
As markers of epistemic access, both evidential and egophoric systems can
have very similar functions when used in interaction. Both types of systems are
used to signal epistemic authority and distribution of knowledge, and both can
participate in the discursive construction of the epistemic stance and status of the
speaker or the addressee.
. Denitions of the relevant interactional concepts
Two concepts used throughout our analysis are epistemic status and epistemic
stance. Epistemic status refers to the knowledge accumulated by the speech-act
participant on the matter at hand. It is inuenced by factors such as relevant per-
sonal experience, the certainty of information, the time when it was acquired, or
the person’s position in the social structure (Heritage b:). Epistemic status is
Egophoricity and evidentiality
relatively stable and does not depend on a communicative situation. The speaker
who has personal experience or more recent information on the topic has a priv-
ileged epistemic status in relation to other participants. Epistemic stance, on the
other hand, is more situation-dependent. The speaker with a given epistemic
status can adopt dierent epistemic stances: either knowledgeable or ignorant,
depending on who they are talking to (e.g., a fellow expert, a layperson), or
their interactional goal (being polite, convincing the interlocutor). Another term
related to epistemic status and stance, and intersecting with them, is epistemic
authority. It can be dened as “the primary right to evaluate the matter assessed”
(Heritage & Raymond :) or the “right to know” (Stivers et al. b:).
Speech-act participants generally treat each other as having primary rights to
talk about their professional and personal experiences (Heritage & Raymond
:).
The concepts discussed above all concern epistemic (i.e., knowledge-related)
rights. Epistemic rights are closely related to epistemic responsibilities, the oblig-
ations of the speakers to have certain information (Stivers et al. b: ). For
instance, it is expected of everyone to know his or her own name, etc. On the
other hand, there is information about other people, their internal states and
experiences, or private aairs, about which their interlocutors do not have a
responsibility, or even a right, to possess knowledge. The domain of epistemic
rights and responsibilities is akin to Kamio’s () ‘Territory of Information’
(henceforth ToI). According to this theory, every person has his or her own terri-
tory of knowledge. The types of information which fall within one’s default ToI are
(i) internal direct experience, (ii) information within one’s professional expertise,
(iii) information obtained through external direct experience including verbal
reports, and considered reliable, (iv) information about persons, object, events
and facts in one’s close environment, and (v) information about oneself (Kamio
:). In terms of domains of knowledge, one has a right to all the informa-
tion listed above, but also a responsibility to be familiar with them. Other types of
information can also become part of one’s ToI when, as Kamio (: –) puts it,
when they are ‘digested and absorbed’ into one’s system of knowledge. The pace
of this process might depend on the type of information – according to Kamio,
more personal information tends to be absorbed sooner.
. Egophoric marking and epistemic authority: The case study of Wutun
This section discusses the interactional functions of egophoric marking in Wutun.
First, the paradigm of egophoric markers is introduced, followed by detailed
examples of their use in spoken interaction. The Wutun egophoric marking has
Erika Sandman and Karolina Grzech
most probably emerged due to language contact with Amdo Tibetan (see
Sandman ), and it shows clear resemblance to egophoric marking systems in
Tibetic languages, which usually contrast ego markers with markers of factual and
sensory information (see e.g., Sun ; Garrett ; Tournadre ; Tournadre
& LaPolla ; Hill & Gawne ; DeLancey ). In Wutun and in Tibetic
languages, egophoric marking does not always mark the actual involvement in the
event under discussion. Instead, expressing personal involvement is oen used as
a tool of negotiating the epistemic rights of the speech act participants. For exam-
ple, the EGO marker can be used metaphorically to give the hearer the impression
that the speaker was present although s/he was not, and therefore the speaker may
use it to claim the primary right to talk about the matter assessed (e.g. DeLancey
). The data discussed in the following sections comes from a h corpus of
Wutun, collected by Erika Sandman in , , and during months of
eldwork in the Qinghai province in P.R. China. The corpus features roughly
speakers of Wutun, and comprises naturalistic interactions, including the ones
analysed in Examples () and (), as well as descriptive and narrative texts and
elicited data, including the Examples ()–().
. The paradigm of egophoric markers in Wutun
Egophoric marking in Wutun is obligatory in all nite clauses, except imperatives
and clauses with prospective aspect marker -zhe, which cannot be combined with
egophoric morphology. Table summarizes the egophoric morphology in Wutun.
Table . Egophoric marking morphology in Wutun
Ego
markers
-yek ego marker, indicates epistemic authority of the speaker
yek ego existential copula
-lio ego perfective
Non-ego
markers
-li . sensory-inferential marker, displays the speaker as an observer of the
event and indicates lack of epistemic authority
re factual evidential auxiliary, is associated with knowledge shared by
the community members
Examples ()–() illustrate the Wutun egophoric marking system. The ego
marker -yek is associated with a high degree of epistemic authority of the speaker,
such as personal involvement in events or states. This explains why an ego marker
is common in rst-person statements (as in ) and second-person questions (as
in ); Speech-act participants generally treat each other as having privileged epis-
temic access to their own experiences, and therefore having primary rights to talk
Egophoricity and evidentiality
about them (Heritage & Raymond :). The sensory-inferential marker -li is
used when the speaker is observing or inferring actions or states of others, as in
() and (), and it is associated with a lower degree of epistemic authority than the
ego marker -yek. Finally, the factual marker re is used when the speaker is talking
about the knowledge shared by the speech-act participants (as in ), and there is
no need to attribute the epistemic authority to the speaker or to the addressee.
() Ego marker in statements
ngu
1
huan
food
xhe-di-yek
drink--
‘I am eating.’ (high degree of personal involvement/epistemic authority)
(Elicited)
() Ego marker in questions
ni
2
ma-ge
what-
nian-di-yek
read--
‘What are you reading?’ (addressee’s personal involvement/epistemic authority)
(Elicited)
() Sensory-inferential marker with second person
ni
2
huan
food
xhe-di-li
drink--.
(Elicited)‘You are eating.’ (as I see/infer)
() Sensory-inferential marker with third person
gu
3
huan
food
xhe-di-li
drink--.
(Elicited)‘S/he is eating.’ (as I see/infer)
() Factual marker
nianha
blind eye
she-wu
ten-ve
tian
day
yek-de
-
re
‘The Losar festival lasts for een days.’ (as we all know)
(Wutun narratives _Festivals)
While the ego and sensory-inferential markers are suxes, the factual marker is
grammatically an auxiliary connected to the preceding verb with the nominalizer
-de. Examples ()–() illustrating Wutun egophoric morphology all come from
elicited sentences or descriptive texts that involve no dialogue between the speech
act participants. Section .. illustrates the use of egophoric marking in naturally
occurring everyday conversation, in which speakers use them to construct their
epistemic authority.
Erika Sandman and Karolina Grzech
. Interactional analysis of Wutun egophoric markers
This section illustrates the interactive uses of Wutun egophoric marking in every-
day conversation, with a special focus on the ego marker -yek, as a way of indexing
epistemic stance. We show that speakers oen use the ego marker -yek in claiming
epistemic authority in conversation. The ego marker can be used in claiming the
primary right to assess the matter at hand or adopting the role of a caregiver in a
particular situation. Conversely, speakers use the sensory-inferential marker -li to
defeat the implicature that they are claiming the primary right epistemic rights in
conversation.
Examples () and () are excerpts from a conversation between four adults.
NZ and her husband DU are a young couple in their twenties, while uncle DH
and aunt ZJ are one generation older and the members of the same extended fam-
ily. In addition, there are two small children involved in the situation: NZ’s and
DU’s -month-old baby boy and an - month-old baby girl, aunt NZ’s grand-
daughter. Example () is uttered when DU is holding the baby girl and uncle DH
is holding NZ’s and DU’s son. The baby boy is cranky and starts crying, and the
adults are discussing that he must be jealous because his father is holding another
baby instead of him.
() DH: je
this
yize
appearance
kan-la1
look-
qhi
vital force
shang-di-li
rise--.
’Looks like (the baby boy) is angry (lit. his vital force is rising).’2
*holding the baby boy------------------------------->*
NZ: jelanba
jealous
gek-ge
dog-
sha-ze
↑hai-yek
-
‘(I am) so jealous.’
DU ¨takes the baby girl LY¨
NZ: £ano: (.)
aba
father
ni
2
kan£
look
‘Oh, father, look!
+points++points+
£ano: (.)
ano: (.)
aba
father
ni
2
kan£
look
‘Oh no, father, look!
. The construction kan-la consists of the verb kan, ‘to see, to look at’ and the conditional
marker -la. It means ‘in view of, looking at’ and is used to express speaker’s observations or
inference. It is also used in forming comparative constructions in Wutun.
. The bodily actions of the speakers are transcribed using conventions developed by Lorenza
Mondada (https://www.lorenzamondada.net/multimodal-transcription).
Egophoricity and evidentiality
DH: ano: (.)
aba
father
ga-mize-ha
little-sister-
bo-she-li-a
hold-.-.-
gu-a
’Oh, the father is holding the little sister.’
( lines omitted)
(the baby boy cries out loud)
NZ: @lanba-ma-da
jealous-.-
qhe-la-li
start--.
’(He) is getting more and more jealous.’
+turns around+
bai-kuu-ge3
-cry-
ya@
‘Don’t let (him) cry.’
DH: gu
3
bai-jhan-ge-da
-see--
be-kuu-yek
-cry-
‘(I) will not let him see, then (he) will not cry.’
*turns around with the baby boy*
*sits on a bench*
guda
there
ga-qhichai
small-car
ga-qhichai
small-car
yek-li
-.
‘There is a toy car over there, a toy car.’
*points at the toy car on the oor*
In line DH proers the rst assessment of the baby boy’s state of mind (je
yize kanla qhi shangdili, ‘looks like he is angry’) while holding the child in his
arms. He uses the construction kan-la, ‘in view of, looking at’ and the sensory-
inferential marker -li in his assessment, showing aliation with the baby boy and
proclaiming that his evaluation of the matter at hand is based on his observations
of the child’s behavior. The mother NZ is responding with a second assessment4
in line , using the ego marker -yek. NZ’s second assessment (jelanba gekge shaze
haiyek, ‘I am so jealous’) can be considered an example of reported speech.5NZ
. The imperative clause here is formed by using the causative marker -ge; in Wutun, third
person imperatives (e.g., Let him see!) are formally causative constructions.
. Following Pomerantz (:), we dene second assessments as assessments that are pro-
duced by recipients of prior assessment, and which have the same referents as prior assess-
ments.
. As in other languages with egophoricity, in Wutun the ego marker -yek in reported clauses
indicates that the person who is the source of the reported information and the person whose
speech is quoted are co-referential, while the sensory-inferential marker -li indicates that they
are not co-referential. Reported speech in Wutun is oen marked by the auxiliary sho, ‘to say,
to speak’ but it is common to omit this auxiliary in naturalistic interactions as in ().
Erika Sandman and Karolina Grzech
is attributing her speech to his -month old son, who still cannot talk himself.
Her response is multifaceted; it voices the baby and projects agreement and al-
iation with DH. At the same time, she claims the primary right to assess her son’s
state of mind by formulating her response like a direct quote from her son and by
using the ego marker -yek, which is a stronger evaluative term than the sensory-
inferential marker -li; indeed, she is the mother. This is in line with some earlier
research on egophoric marking in child-caregiver interactions, which has demon-
strated that in such interactions the mother or a caregiver oen adopts a role of
primary epistemic authority in relation to the child’s actions (see San Roque &
Schieelin :). NZ’s response to DH can be considered as an example of
upgraded second assessment (see Pomerantz :); it incorporates stronger
evaluative term than the prior assessment (the ego marker -yek instead of sensory-
inferential marker -li). While NZ is uttering her response in line , the father DU
is taking the baby girl in line . NZ continues with two direct quotes attributed
to his son in lines and while pointing at DU, who is holding the baby girl.
The use of reported speech, as well as the interjection anoo which has an aec-
tive meaning, imply strong aliation with her son. DH responds with a state-
ment (ano aba gamizeha boshelia gua, ‘oh, the father is holding the little sister’) in
lines , using the sensory-inferential marker -li. His statement with the sensory-
inferential marker -li shows aliation with the baby boy and, but by using -li he
denies the privileged access to information necessary for claiming primary right
over the mother NZ to evaluate the baby boy’s behaviour.
Lines – illustrate how the ego marker -yek is used in shiing from the
role of an observer to a role of a caregiver. In line the baby boy becomes even
more restless and starts to cry out loud. Mother NZ reacts to her son’s crying by
requesting DH and DU do something to calm the baby in lines –. In line
DH responds to NZ’s request by turning around with the baby boy and prevent-
ing him from seeing his father DU holding the baby girl, as well as responding
with the statement gu baijhangeda bekuuyek, ‘I will not let him see (them), then
he will not cry.’ While uttering his response, DH is taking the baby boy physically
closer to him and turning him around, so his bodily behaviour is aligning with his
assertion of authority and addressing the concern of the mother. Finally, DH sits
down with the baby boy and starts pointing to the toy cars on the oor to catch
the child’s attention in lines and . We have seen that DH, who has only used
the sensory-inferential marker -li in commenting on the baby boy’s behaviour in
lines –, suddenly switches to ego marker -yek in line . This can be attrib-
uted to the fact that in line he takes the role of a caregiver in the situation. The
role of a caregiver is intimately linked to epistemic authority; the caregiver is in
charge of decisions that aect the person in their care, authorizing that person to
assess the other’s state of mind (see Section ).
Egophoricity and evidentiality
In Example (), ego marker -yek occurs in a context where the speaker is seek-
ing a more vigorous alignment with his directive and warning another person.
The Example () is uttered when the aunt ZJ prepares to leave the house with her
-month old granddaughter and DH instructs her to put enough warm clothes on
the child.
() DH: da
now
qhi-zhe
go-
ma
’Are you going now?’
ZJ: da
now
qhi-zhe
qo-
’We are going now.’
DH: ya
da
now
ni
2
qhi
go
’Ok, you go now.
quandi
clothes
quan-she-ma
put on-.-
rai~rai-de
warm~warm-
ze-she
do-.
‘Put some clothes (on her) so it will be warm.’
ZJ: ya
ok
DH: nanqhan
u
lai-gu-yek
come--
’Otherwise she will get a u.’
*coming close to the baby and holding her hand*
In lines –, aunt ZJ is preparing to leave the house and take the baby girl with
her. Uncle DH reacts by asking about her plans to leave in line and she conrms
them in line . Then in lines – DH proceeds with a directive and instructs
aunt ZJ to put some warm clothes on the child. Aunt ZJ responds with the ar-
mative interjection ya, ‘ok’ in line . DH treats this as a weak agreement and
responds with a warning ‘She will get a u’ (literally: ‘Flu is coming’) in line .
He uses the ego marker -yek in his warning, which can be interpreted as seeking
a more vigorous alignment with his directive and adopting a knowing epistemic
stance.
Erika Sandman and Karolina Grzech
. ‘Evidentials’ as markers of epistemic authority: The case study of
Upper Napo Kichwa
UNK has a rich epistemic paradigm, presented in Table .
Table . Epistemic enclitics attested in Upper Napo Kichwa
marker meaning
=mi speaker-exclusive epistemic authority and access; proposition likely to be unexpected to
the addressee; contrastive focus
=ma similar to =mi
=mari authority/access shared between speaker and addressee, addressee acting as if unaware
of this
=chu polar question/negation marker
=cha disclaimer of epistemic authority
=chari similar to =cha
=tá speaker’s epistemic authority, access shared between speaker and addressee; ‘verum’
focus6
=ta information question marker
The analysis in this section focuses, above all, on the enclitic =mi, which is the
cognate of direct evidentials in other Quechuan varieties (Floyd ; Faller ;
Hintz & Hintz ; cf. Grzech a). In UNK, =mi encodes exclusive epistemic
authority of the speaker, and is associated with (contrastive) focus (Grzech ;
a; c). The analysed data comes from a h corpus (Grzech b), col-
lected in – during months of eldwork. The corpus contains data from
over speakers, and comprises naturalistic interactions, as well as interactive
elicited discourse based on dierent types of stimuli.
. Meaning and interactional functions of UNK evidentials/epistemic
markers
Like evidentials in other Quechuan varieties, UNK epistemic clitics occur on
phrasal heads, and are not associated with a particular lexical category. Unlike
Wutun egophoric markers, UNK ‘evidentials’ are not obligatory for the gram-
. While other markers listed in Table are attested in other Quechuan varieties, it is not
clear whether =tá has cognates in Quechuan varieties described to date (cf. Grzech : ,
–).
Egophoricity and evidentiality
maticality of utterances. The markers =mi and =cha are cognates of the direct
and conjectural evidentials, respectively, in other Quechuan languages. However,
source of information is relevant to their meaning only indirectly (see Table ).
They signal epistemic authority or lack thereof, as well as distinctions between
individual and shared knowledge. Epistemic markers are used in interaction to
construct epistemic stance and status (Heritage a, see Section ), presenting
certain information as pertaining – or not – to the ToI of the speaker (Kamio ,
see Section ) and thus conveying the speaker’s claims to epistemic authority (e.g.,
Heritage & Raymond ; Stivers et al. b).
Examples () and () illustrate the complex epistemic distinctions encoded by
the UNK markers. The participants are the same in both examples: KG: Karolina
Grzech, and B: KG’s friend and a native speaker of Kichwa. B is an expert in
brewing chicha, a traditional drink made of manioc, whereas KG has never made
chicha before. Example () is extracted from a spontaneous interaction, which
occurred naturally. Example () was elicited: The native speaker was asked to
encourage a novice to keep making an eort in a context analogous to that of ().
() A is making chicha for the rst time, but does not think she is up to the task.
KG: Mana
usha-ni.
can-1
‘I can’t [make chicha].’
B: [Kan]
[2]
usha-ngui
can -2
=mi!
=
‘[Yes, you] can!’
() KG is attempting to make chicha for the rst time. B wants to encourage her to
keep up the eort.
B: [Kan]
[2]
usha-ngui
can -2
=rá!
=
‘[You] can [do it]!’
In (), B makes an encouraging, =mi-marked statement aer KG says she will not
manage to make chicha. B is the expert chicha-brewer, so epistemic authority lies
with her. Because KG just said she will not succeed, B also has reasons to believe
that her utterance is unexpected to KG. This combination of speaker-exclusive
epistemic authority with speaker-exclusive access to the proposition warrants B’s
use of =mi in (). In (), B’s expert status remains unchanged. However, KG does
not voice concerns regarding her ability to make chicha, and is actively engaged in
the task. Thus, B can assume that KG is expecting to succeed. This combination
of B’s epistemic authority and KG and B’s shared access to the proposition war-
rant the use of =tá in Example ().
Erika Sandman and Karolina Grzech
Example () comes from a conversation between two women, SA and LC.
They belong to the same extended family – SA is married to LC’s nephew –
and live within a few hundred metres of one another. The conversation, which
took place when they were working on a land plot,7is divided into shorter
stretches, presented in Examples () through (). Line numbers continue across
the examples to give a sense of how the original conversation developed.
() LC: J.
{compadre}8,9
compadre
ima
what
wasi-ra
house-
ra-w-n?
make--3
‘What building is compadre J. going to build?’
{Cosina}?
kitchen
‘[The] kitchen?’
SA: Mana,
mana
ra-w-n=mari
make--3=
‘No, he’s not building…’
LC: Tsatsa-ra,
sand-
wasi
house
ra-nga=cha,
make-=
asta-nga
load-
ra-w-n
--3
‘The sand, to make the house, [he] will carry [it]…’
In lines and LC asks about the actions of J.: LC’s nephew and SA’s husband.
LC assumes that J. is renovating his house, and particularly the kitchen. LC’s
assumes this, because she knows J. borrowed a wheelbarrow from her father (see
line in () below). In line , SA contradicts LC’s assumptions, stating that her
husband is not building anything. SA uses the epistemic marker =mari, which sig-
nals that a given piece of information should be known to the addressee, although
the addressee acts/speaks as if it was not (Grzech , see Table ). By using
=mari, SA reinforces the statement in line . In line , LC’s reaction corrob-
orates the semantics of =mari explained above. SA’s use of =mari indicates LC
should know that no building is happening, so LC justies her assumption by
referring to an event she observed: J. came to her house to borrow the wheel-
barrow. The utterance in line is a complex sentence. LC claims that J. will
bring sand – based on the fact that he came to borrow the wheelbarrow (tsatsara
. This recording and the accompanied les are deposited in the ELAR archive (Grzech
b). Recording code: ev__, session title: ‘Working in the chagra/Trabajo en la
chagra’.
. Following Nikolaeva (2014), words in the regional lingua franca (Spanish) are enclosed in
{}.
. Compadres are godparent’s to each other’s children, like family to one another. The cultural
practice of compadrazgo is extremely important among UNK speakers.
Egophoricity and evidentiality
astanga rawn, ‘he will be bringing sand’). This statement is not epistemically qual-
ied. However, in the same line LC also says that J. will use the sand to build the
house (wasi rangacha, ‘to make the house=’). This statement is marked with
=cha, signalling the speaker’s lack of epistemic authority (Grzech ; c, see
Table ): likely a reaction to SA’s claim from the previous turn (that J. is not build-
ing anything). The conversation goes on as shown in ().
() SA: Aj,
chi-pa-y
.--
moto-y
motorbike-
puri-sha
go-
ni-kpi
say-
‘Ah, [J.] said it’s no good to pass there on the motorbike…
mana
{bali}-jpi,
be.good- .
chi
drop
sanka ni-n
say-3
yachi-n,
seem-3
I think (lit. it seems) he said there are holes [in the path],
wayku
fall
mana
{bali}-n
be.good-3
ni-n……
say-3
[with] holes it is no good, he says.’
LC Mjm
‘Yeah…’
SA Chi-bi=mi,
.-=
ishki
two
{caretilla}-ra=mi
wheelbarrow-=
tsatsa-ra
sand-
api-sha
bring-
‘There, two wheelbarrows, bringing the sand,
shita-nga
throw-
ra-w-n
--3
ni-sha, maña-chi-nga
say- ask--
kacha-w
send-
yachi-n.
seem-3
saying [he] will throw it there, [he] sent (our son) for (the
wheelbarrow), it seems.’
Aer LC’s admission of lack of epistemic authority in line of Example (), cou-
pled with her justication of the source of information (she saw J. borrowing
the wheelbarrow), SA realizes the reason for the misunderstanding. This can be
inferred from the beginning of line in Example (), where SA uses a ‘change
of state token’ (cf. Heritage ): the interjection ‘ah.’ In lines , and , SA
oers a detailed explanation: J. needs the sand not to build, but to even up the
path leading to their house. SA reports her husband’s claim about the path being
uneven, using the verb nin ‘[he] says’. Moreover, SA reduces her own epistemic
authority by using the epistemic verb yachin ‘[it] seems’ in line . These strate-
gies are instances of epistemic downgrading. In this context, SA has primary epis-
temic authority, since she speaks about her house and her husband. However,
socially she has a lesser epistemic status than LC, who is more ‘senior’ in the fam-
ily hierarchy. Thus, SA chooses not to exploit her privileged epistemic rights. In
Erika Sandman and Karolina Grzech
line , LC accepts the explanation, but does not oer any new contribution. This
prompts SA to oer an additional explanation in lines and . Here, she uses
the epistemic clitic =mi twice: rst with the demonstrative which prefaces the
turn, and then when reporting a claim made by her husband. The second occur-
rence of =mi in line marks the report of the husband’s original utterance, and
hence the epistemic authority expressed by =mi is indexed to him, and not to SA.
Moreover, when SA oers her own observation in line , she downgrades her
epistemic stance with the verb yachin ‘[it] seems’, in what could be a continued
eort to attend to LC’s superior epistemic status as a senior family member. Given
this explanation, LC needs to review her idea about J.’s plans. Further conversa-
tion is presented in Example ().
() LC Mjm
‘Yeah’
Sarpundza
day.before.yesterday
maña-n
ask-3
tsatsa-ra=mi
sand-=
astanga
carry-
ra-w-n
--3
‘The day before yesterday he borrowed (the wheelbarrow), he will
bring the sand,
shina
like.this
ni-ka,
say-3.
wasi-ra=chu
house-=/
ra-nga
make-
ra-w-n
--3
he said, will he make a house
ima=chari,
what=
“{cosina}
kitchen
wasi-ra=mi…
house-=
or what, “the kitchen building…
( line omitted, interruption from another speaker)
… ra-nga
make-
ra-w-nchi” ni-n,
--1
kacha-w-ka
say-3
ni-n.
send--3. say-3
we will build” (he) was going around saying.’
SA Mjm
‘Yeah’
LC Chi-ta
.-
ra-nga
make-
minishti-shka a-nga,
need- -
shina
like.this
ni-w-ka
say--
‘[It is] to do this [that] he will be needing [the wheelbarrow], he was
saying this.’
SA Mana
‘No’
Egophoricity and evidentiality
In line , LC accepts SA’s previous contribution with an interjection. In lines
and , LC recognizes she was mistaken, and uses epistemic markers and other
stance-constructing strategies to justify her prior position. In line , LC reports
what J. said, marking the report with =mi. Immediately aer, she oers her
own statement, marked with the polar question/negation marker =chu (line ),
and the marker =chari, which, like =cha, is a disclaimer of epistemic authority
(line ). Thus, LC re-frames the statement about J. building something as a
supposition, deferring her own epistemic authority over the interpretation of J.’s
words. In line and , LC provides another quote from J., claiming that he said
he would build the house. By quoting J. and constructing him as the authority, LC
frees herself of epistemic responsibility for the statement which turned out to be
problematic in prior conversation (that J. is building something). Whether or not
LC has really heard from J. that he was going to build the house is uncertain – the
transcriber, who happened to be J. himself, observed that LC did not have the cor-
rect information. In line , SA responds with just an interjection. LC follows in
line , re-conrming what she said previously, again with a report of J.’s words. In
line , SA responds with the negative particle mana. This is a direct disagreement
with LC’s previous turns, but, since they were framed as reports and quotations,
direct negation is not face-threatening – it does not target the epistemic status of
LC, as she cannot be held responsible for the veracity of the reported statements.
Throughout the conversation, the construction of stance through the use of
epistemic clitics intersects with other tools, such as introducing reported dis-
course. The interlocutors’ use of linguistic strategies show their awareness of each
other’s status of the interlocutors, which can be derived from a variety of sources:
age, family seniority, proximity to the talked-about events, etc. In the society of
intimates, of which the UNK speaker community is an example, there can be mul-
tiple sources of epistemic rights and responsibilities relative to one piece of infor-
mation (Mushin ).
Excerpt (), given below, immediately follows (), hence the continuing
numbering of the lines. In this example, epistemic authority is tied specically to
adopting a particular discursive role: that of a caregiver. The new topic of conver-
sation is Grzech: the linguist who arrived in the community ten days previously,
and who is staying at LC’s house. LC is now talking to YC, who is her peer in the
family structure, as they are of the same generation. YC has oered Grzech the
chicha, which prompted the following exchange:
Erika Sandman and Karolina Grzech
() LC [whispers] Apama
woman
upi-shita-n=lla=ma
drink-throw-3==
asa-ra.
chicha-
‘That girl, how she just throws chicha in herself ’
YC [in Spanish] Tome, tome, tiene que emborrachar…
‘Drink, drink, you have to get drunk…’
LC [whispers] Mana,
macha-nga=mi,
get.drunk-=
Y.!
‘No, Y., she’ll get drunk!’
( lines omitted)
LC [whispers] Mana
kikin-da
our.own-
upi-n=dá,
drink-3=
upi
drink
upi,
drink
shinallara
like.this
mana
ashka-ra
much-
upi-n.
drink-3
‘She does not drink like a local, she drinks a little, but not much.’
( lines omitted)
In line , LC makes a comment about Grzech’s drinking capacities, marked with
=ma, most likely indexing the speaker’s epistemic authority (see Table ). LC
whispers, which conrms that she sees her statement as infringing on Grzech’s ToI
by evaluating her actions. In line , YC exploits LC’s worry and Grzech’s new-
ness to the community for the purpose of a joke: she tells Grzech in Spanish that
she should get drunk, which, in fact, is not socially acceptable. Line shows that
LC suspects Grzech will take YC’s joke at face value. LC whispers again, because
she is infringing on Grzech’s ToI by evaluating her bodily capacity, but she also
uses =mi, thus claiming epistemic primacy with respect to what she is assessing.
Assuming epistemic primacy with respect to somebody else’s internal states or
future actions is not customary. Thus, it can be analysed as indicating that LC
assumes a caregiver position with respect to Grzech, treating her as is she was a
child (see Section . on Wutun). This parallels the use of =mi in speech acts of
admonition or warning, usually issued from the same caregiver position. That LC
does assume a caregiver role with respect to Grzech becomes more evident as the
conversation develops. In line , LC reasons with YC, using the epistemic marker
=tá, which indexes the epistemic authority of the speaker, but also acknowledges
shared epistemic access (see Table ). YC also got to know Grzech as soon as she
arrived in the village. LC now explicitly draws on their shared experience, while
maintaining epistemic primacy as Grzech’s ‘host mother.’ The latter is evidenced
through the choice of =tá over =mari, which would have indexed YC’s responsi-
bility to know about the Grzech’s chicha-drinking capabilities (see Example ()).
That LC does wish to adopt a caregiver role is further evidenced in how the
conversation develops in Example ().
Egophoricity and evidentiality
() LC [whispers] Macha-nga=mi
get.drunk-=
‘[She]’ll get drunk’
Sarpundza
day.before.yesterday
angu
sand.mite
llaki
lluta-ri-shka
stick--
apama-ra
woman-
‘The day before yesterday, loads of sand mites stuck to the girl!’
( lines omitted)
LC Shina
like.this
llushpi-chi-sha
peel--
riku-chi-ni.
see--1
‘Like this, peeling them o, I showed (her).’
In line LC whispers again, while also reinforcing her =mi-marked evaluation
of Grzech’s capacity to drink. In the following lines, LC no longer whispers or
uses epistemic marking when recounting how she helps Grzech adjust to the local
life. This signals that LC views this content as more neutral, possibly because
it does not infringe on Grzech’s capacity to evaluate her own actions. In these
nal lines, LC reinforces her caregiver role not through explicit claims of epis-
temic authority, but through supplying additional evidence of how she looks aer
Grzech, preforming actions characteristic of a caregiver. This shows that discur-
sively constructing oneself as a caregiver can be achieved through a combination
of strategies. Similarly to what we have seen for Wutun, in UNK epistemic mark-
ing features prominently among these strategies.
. Discussion
In the previous sections, we have highlighted the functional similarities between
Wutun and UNK epistemic systems in interaction. Similar interactional uses sug-
gest, in line with observations made in the literature, that epistemic systems are
exible: How they are used depends on the epistemic stance speakers want to pro-
ject, rather than on the discourse-external reality (cf. e.g., Mushin ; Hintz &
Hintz ; Sun ). The discussion in this section focuses on this exibility
with particular reference to the Wutun ego marker -yek and the UNK ‘direct evi-
dential’ =mi.
As mentioned in Section , literature on epistemicity oen assumes that more
grammaticalized epistemic systems are less exible in terms of encoding mean-
ings which are – from the theoretical standpoint – not at the core of their epis-
temic semantics. However, the Wutun data discussed in Section clearly shows
that even though egophoricity in Wutun is obligatory for the grammaticality of
utterances, it can still be used exibly, not to encode speaker involvement, but
Erika Sandman and Karolina Grzech
rather epistemic authority and the knowing or privileged epistemic stance of the
speaker. This is shown acutely in Example () in Section ., where the same
speaker (DH) rst talks about the same situation with the sensory-inferential
marker -li, and then, once he assumes a dierent role in the interaction, switches
to the EGO -yek. For UNK ‘evidentials’, this exibility is shown in Examples ()
and () in Section ., where the marker is chosen depending on the speaker’s
assumption about the addressee’s knowledge, and not because of the source of
evidence for the conveyed proposition. Sections and show that the Wutun
-yek and the UNK =mi full similar interactional functions: They are used when
the speaker wishes to assert epistemic authority and project a knowing epistemic
stance. Similar observations have been made by many authors who acknowledge
that speakers use evidentials depending on their interactional goals, e.g., to nego-
tiate responsibility for events (cf. e.g., Michael for Nanti) or ownership of
knowledge (cf. e.g., Hintz & Hintz for varieties of Peruvian Quechua).
The exibility discussed above is shown acutely for both Wutun and UNK
in how their speakers use epistemic markers when adopting the role of caregiver.
As a caregiver, one can make claims regarding the Territory of Information (see
Section ) of the people one looks aer, as if the ToI of the guardian encompassed
the ToI of their ward. When acting as caregivers, we assume primary access to
information concerning others, and treat the knowledge regarding our ward as
if it was self-knowledge, even though in other discursive contexts we distinguish
between self-knowledge and knowledge of others, given that they are acquired in
fundamentally dierent way (cf. Garrett , cited in Norclie :).
In Wutun Examples () and (), when adopting the caregiver role, speakers
choose to use the EGO marker -yek, even if the degree of involvement in the
described action does not license it. The same is true for UNK Examples ()
and (), where the speaker uses the ‘direct evidential’ =mi when talking about
the linguist, who, as a novice in the community, has a child-like status in many
respects. In UNK conversation between two equals, it would not be felicitous for
one speaker to talk about the other’s actions, thoughts, or intentions, using =mi.
Thus, the Wutun and UNK data can be treated as parallel, even though the con-
versations occurred in dierent interactional contexts.
It can safely be assumed that the role of caregiver is both socially and discur-
sively universal. The prototypical example of a caregiver role, which occurs across
all human societies, is someone taking care of an infant: a parent, grandparent,
older sibling, etc.. The boundaries of who can assume such a role vary across
communities, and dierent societies allow people in other social roles – elders,
leaders, preachers, teachers, etc. – to discursively adopt a caregiver-like role by
assuming superior epistemic rights. Despite these dierences, caregiver can be
considered a universal social and discursive macro-role, involving a family of dis-
Egophoricity and evidentiality
cursive actions (cf. Sacks ), such as issuing warnings, giving advice, or speak-
ing on behalf of others. We are not aware of cross-linguistic studies of warnings
and advice-giving, but given the social universality of the caregiver role, it can be
expected that other languages allow for non-prototypical uses of epistemic mark-
ing in such contexts. In fact this is the case for Japanese, where the particle yo,
indexing epistemic primacy (e.g., Hayano ), can be used in these situations,
or for Mangghuer (Amdo Sprachbund, China, Fried ), where an ego (‘subjec-
tive’) copula is used to express a high degree of concern for the interlocutor, or
when speaking to those with lower epistemic status (Fried :, ).
We do not mean to posit any categorical claims regarding the relationship
between categories of evidentiality and egophoricity, as attested cross-
linguistically. We merely bring together interactional analyses of data from two
languages, spoken in the two hotbeds of grammaticalized epistemic marking:
South American and the Greater Himalayan region. Our data suggests that the
epistemic systems from these two areas share many functional similarities, which
have to date been obscured by two dierent research traditions. Research on lan-
guages of the Greater Himalayan region has given us the earliest descriptions of
egophoricity (Hale ; Bendix ), and many studies on epistemic marking
in the Himalayan and Tibeto-Burmanist tradition still focus on the relationship
between evidentiality and egophoricity (cf. e.g., Tournadre ; Widmer ;
Zemp ). While recent research has also focused on pragmatic meanings and
interactional uses of epistemic markers (cf. e.g., Tournadre & LaPolla ; Zeisler
; DeLancey ), conversation analytic studies on epistemic marking in the
Himalayan languages are still virtually non-existent.
In South America, epistemic marking is largely analysed within the evidential
framework inuenced by the tenets proposed by Aikhenvald (), but in the
recent years the interactional uses of the epistemic markers have also been
observed and analysed (cf. e.g., Michael ; Gipper ; Nuckolls & Michael
; Bergqvist ; Hintz & Hintz ; Grzech ). Egophoricity is not
completely unattested in South America. It has been described for example in
the Barbacoan language family, spoken in Colombia and Ecuador. The family
encompasses six languages, three of which – Tsaki, Guambiano/Nam Trik, and
Cha’palaa – exhibit both evidential and egophoric marking. Tsaki (Dickinson
; p.c.) evidential and egophoric markers do not occur in paradigmatic dis-
tribution. Guambiano/Nam Trik has epistemic markers indexing both speaker
involvement and sources of evidence. Unlike in Tsaki, in Nam-Trik these two
types of markers form one paradigm, which Norclie () analyses as indicating
evidence type (visual -tre vs. non-visual -shi) and evidence strength (-an). In
Cha’palaa, egophoric and evidential markers also occur in paradigmatic distrib-
ution (Floyd ). This observation shows that the relationship between mark-
Erika Sandman and Karolina Grzech
ers analysed as egophoric and evidential can dier even between closely related
languages, and provides additional support for our argument, namely that to
better understand the relationship between egophoricity and evidentiality, we
should look at the function of these marking systems, rather than assuming that
by comparing morphemes which were assigned the same labels we are compar-
ing like with like. This speaks to the need to develop comparative concepts (cf.
Haspelmath ) also in the more context-sensitive branches of linguistics, such
as pragmatics, or Conversation Analysis.
Therefore, this paper also contributes to the incipient enterprise of pragmatic
typology (cf. Dingemanse & Floyd ; Dingemanse, Blythe & Dirksmeyer
; Kendrick et al. ; Floyd ). Thus far, pragmatic typology is mostly
done with a ‘natural control method’ (Stivers et al. ), identifying similar
structures across languages, or with a ‘sequential control method’ (Dingemanse
& Floyd ), which consists in identifying particular types of interactional
sequences. These methods are employed for cross-linguistic comparisons of dis-
cursive actions such as expressions of gratitude or recruitments (cf. Floyd et al.
, contributions to Floyd et al. ). Our ndings suggest that discursive
macro-roles adopted by speakers are a potentially important parameter, which
could be factored into the pragmatic-typological approach alongside structures or
sequence types. Discursive macro-roles can be associated with actions, or fami-
lies of actions, and these, in turn, are represented by particular discourse units,
or sequences of units. Thus, an approach incorporating discursive macro-roles
introduces an additional functionally-motivated and intersubjective layer into the
pragmatic-typological picture.
A comment is also in order with respect to the importance of interactional
linguistics (IL) as an approach for investigating pragmatics comparatively and
typologically. Many recent studies on epistemicity, despite using naturalistic data,
merely acknowledged the context in which these data occurred (cf. e.g., Grzech
). A thorough functional analysis of epistemic marking needs to go beyond
this, towards an in-depth and systematic analysis of the principles that shape the
interactional context. IL methods are an ideal tool for achieving this goal, and they
are increasingly being recognized as applicable not only to the mother tongues
of researchers doing IL, but also to minority and endangered languages, even if
research on them is done by outsiders (cf. Mushin ).10 By giving researchers
tools to systematically describe complex aspects of interaction, it allows for trian-
gulation of results. This, in turn, opens the possibility for comparative research
that truly goes beyond dierent descriptive research traditions or analytical labels.
. Consider, e.g., the CIARA project, researching conversational style in both Australian Eng-
lish and Australian Aboriginal languages (https://www.ciaraproject.com/, PI: Joe Blythe).
Egophoricity and evidentiality
. Conclusions
This paper contributes to the studies of epistemicity in several respects. Firstly, it
adds to descriptive studies of epistemic systems, and makes an argument for the
need to analyze such expressions in their interactional context. The paper shows
that even grammaticalized epistemic marking systems are used in ways which
depend primarily on the interpersonal conguration of the interaction, including
epistemic stance and discursive roles which speakers wish to project. Secondly,
the paper puts forward new data from two under-described languages, adding
to the studies of epistemicity within an interactional linguistic framework. Our
analysis brings out similarities in the use of egophoric marking in Wutun and evi-
dential/epistemic marking in UNK which can only be appreciated from an IL per-
spective, and thus makes the case of analysing more data on epistemic systems in
lesser-described languages with interactional linguistic methods.
Furthermore, we believe that the functional parallels between Wutun
egophoric marking and UNK evidential/epistemic marking that we describe in
the paper have both methodological and theoretical implications. In terms of
methodology, our study demonstrates the importance of documenting natural
interaction and the use of epistemic marking within it. In applying IL method and
concepts to lesser-documented languages exhibiting grammaticalized epistemic
systems, we show that they can be compared with better-described languages in
terms of how they make use of stance-taking. Explorations of this type could con-
tribute new and valuable information to IL studies. In terms of theory, our paper
strengthens and substantiates the claims that egophoric and evidential marking
can full similar discourse functions. Moreover, our analysis shows that the inter-
actional functions of evidential and egophoric marking can have more to do with
epistemic status, stance, and authority, than with involvement or source of infor-
mation.
Funding
We would like to thank our funders: the KONE Foundation (Erika Sandman) and the Endan-
gered Languages Documentation Programme (Karolina Grzech, grant no. IGS).
Acknowledgements
We are also grateful to the Wutun and Upper Napo Kichwa communities who hosted us for
our respective research. This paper beneted greatly from our discussions with many people.
Among them are the members of the former Centre of Excellence ‘Intersubjectivity in Interac-
tion’ at the University of Helsinki, especially Marja-Leena Sorjonen, Auli Hakulinen, Elizabeth
Erika Sandman and Karolina Grzech
Couper-Kuhlen, Jan Lindström, Salla Kurhila, Yida Cai, Suvi Kaikkonen, Katariina Harjunpää.
The ideas in the paper also beneted from the insights of the members of the General Linguis-
tics research group at the Linguistics Department at Stockholm University, especially from our
discussions with Henrik Bergqvist. We would also like to thank Simeon Floyd and an anony-
mous reviewer for insightful comments on the earlier dra of the paper. Finally, we would like
to thank Ilana Mushin. Her ideas inspired much of the thinking in this paper, and her editorial
work enabled its swi publication.
Symbols used in transcripts
(.) brief interval, between . and . seconds
Word emphasis
Wo::rd prolonged vowel or consonant
↑rising pitch
smiley voice
@ other change in voice quality
Abbreviations
rst person
second person
third person
accusative
anterior
anticausative
augmentative
auxiliary
causative
coordinative
comitative
completive
conditional
consequential
copula
co-reference
. distal demonstrative
ego
emphatic
equative
existential
factual
future
genitive
incompletive
interjection
limitative
locative
proper name
negative
nominalizer
plural
progressive
past
progressive
prohibitive
particle
purpose
question
referential
. agent-oriented resultative
. patient-oriented resultative
. sensory-inferential
singular
switch-reference
Egophoricity and evidentiality
References
Aikhenvald, A.Y. (). Evidentiality. Oxford: OUP.
Aikhenvald, A.Y. (Ed.), (). The Oxford Handbook of Evidentiality. vol. . Oxford, New
York: OUP. https://doi.org/10.1093/oxfordhb/9780198759515.013.2
Aikhenvald, A.Y. & LaPolla, R. J. (). New perspectives on evidentials: A view from Tibeto-
Burman. Linguistics of the Tibeto-Burman Area (). –.
Aikhenvald, A.Y. & Dixon, R.M. W. (Eds.), (). The Grammar of Knowledge: A Cross-
Linguistic Typology. Explorations in Linguistic Typology . Oxford, New York: OUP.
https://doi.org/10.1093/acprof:oso/9780198701316.001.0001
Bendix, E. (). The grammaticalization of responsibility and evidence: Interactional
potential of evidential categories in Newari. In J.H. Hill & J.T. Irvine (Eds.) Responsibility
and Evidence in Oral Discourse (pp. –). Cambridge: CUP.
Bergqvist, H. (). Complex epistemic perspective in Kogi (Arawako-Chibchan).
International Journal of American Linguistics (), –. https://doi.org/10.1086/684422
Bergqvist, H. (). The role of ‘perspective’ in epistemic marking. Lingua –, –.
https://doi.org/10.1016/j.lingua.2015.02.008
Bergqvist, H. & Grzech, K . In Press. The role of pragmatics in the denition of evidentiality.
STUF – Language Typology and Universals.
Bergqvist, H. & Kittilä, S. (). Evidentiality, egophoricity and engagement. Berlin: Language
Science Press. https://doi.org/10.5281/ZENODO.3968344
Bergqvist, H. & Knuchel, D. (). Complexity in egophoric marking: From agents to attitude
holders. Open Linguistics , –. https://doi.org/10.1515/opli-2017-0018
Cli, R. (). Indexing stance: Reported speech as an interactional evidential. Journal of
Sociolinguistics (), –. https://doi.org/10.1111/j.1467-9841.2006.00296.x
Cornillie, B. (). Evidentiality and epistemic modality: On the close relationship between
two dierent categories. Functions of Language (), –.
https://doi.org/10.1075/fol.16.1.04cor
Couper-Kuhlen, E. & Selting, M. (). Interactional Linguistics: An Introduction to Language
in Social Interaction. Cambridge: CUP. https://doi.org/10.1017/9781139507318
DeLancey, S. (). Evidentiality in Tibetic. In A. Y. Aikhenvald (Ed.) The Oxford Handbook
of Evidentiality Online.https://doi.org/10.1093/oxfordhb/9780198759515.013.27
Dickinson, C. (). Mirativity in Tsaki. Studies in Language (), –.
https://doi.org/10.1075/sl.24.2.06dic
Dingemanse, M. & Floyd, S. (). Conversation across cultures. In N.J. Eneld,
P. Kockelman & J. Sidnell (Eds). Cambridge Handbook of Linguistic Anthropology.
Cambridge: CUP. https://doi.org/10.1017/CBO9781139342872.021
Dingemanse, M., Blythe, J., & Dirksmeyer, T. (). Formats for other-initiation of repair
across languages: An exercise in pragmatic typology. Studies in Language (), –.
https://doi.org/10.1075/sl.38.1.01din
Faller, M.T. (). Semantics and Pragmatics of Evidentials in Cuzco Quechua. PhD Thesis.
Stanford University.
Floyd, R. (). La estructura categorial de los evidenciales en el quechua wanka. Serie
Lingüística Peruana . Lima: SIL International.
Erika Sandman and Karolina Grzech
Floyd, S. (). Chapter . Egophoricity and argument structure in Cha’palaa. In S. Floyd,
E. Norclie & L. San Roque (Eds.) Egophoricity. Typological Studies in Language (pp.
–). Amsterdam: John Benjamins. https://doi.org/10.1075/tsl.118.09flo
Floyd, S. (). Conversation and Culture. Annual Review of Anthropology (), –.
https://doi.org/10.1146/annurev-anthro-101819-110158
Floyd, S., Rossi, G., Baranova, J., Blythe, J., Dingemanse, M., Kendrick, K. H., Zinken, J. &
Eneld, N.J. (). Universals and cultural diversity in the expression of gratitude. Royal
Society Open Science (), . https://doi.org/10.1098/rsos.180391
Floyd, S., Rossi, G., Eneld, N. J., Kendrick, K . H., Blythe, J., Zinken, J., Baranova, J. &
Dingemanse, M. (). Getting others to do things. Berlin: Language Science Press.
https://doi.org/CitetononCRdoi:10.5281/zenodo.4017493
Fried, R. W. (). Chapter . Egophoricity in Mangghuer: Insights from pragmatic uses of the
subjective/objective distinction. In S. Floyd, E. Norclie & L. San Roque (Eds.),
Egophoricity. Typological Studies in Language, vol. (pp. –). Amsterdam: John
Benjamins. https://doi.org/10.1075/tsl.118.07fri
Garrett, E. J. (). Evidentiality and Assertion in Tibetan. Ph.D. Thesis. Los Angeles:
University of California.
Gipper, S. (). Evidentiality and Intersubjectivity in Yurakaré: an Interactional Account.
PhD Thesis. Nijmegen: MPI.
Gipper, S. (). Intersubjective evidentials in Yurakaré: Evidence from conversational data
and a rst step toward a comparative perspective. Studies in Language (), –.
https://doi.org/10.1075/sl.38.4.05gip
Gipper, S. (). (Inter)subjectivity in interaction: Investigating (inter)subjective meanings in
Yurakaré conversational data. STUF – Language Typology and Universals (), –.
https://doi.org/10.1515/stuf-2015-0011
Gipper, S. (). Conversational structure as evidence for regularity and variability in the use
of epistemic markers: The case of Yurakaré. Presentation given at the nd Annual
Meeting of the Societas Linguistica Europaea, Leipzig.
Grzech, K. (). Discourse enclitics in Tena Kichwa: A corpus-based account of information
structure and epistemic meaning. PhD Thesis. SOAS, University of London. https://
eprints.soas.ac.uk// (access //).
Grzech, K. (b). Autoridad epistémica y atenuación en Tena Kichwa: Análisis de enclítico
=cha basado en el corpus. Normas (), –. https://doi.org/10.7203/Normas.7.11167
Grzech, K. (a). Managing Common Ground with epistemic marking: ‘Evidential’ markers
in Upper Napo Kichwa and their functions in interaction. Journal of Pragmatics ,
–. https://doi.org/10.1016/j.pragma.2020.05.013
Grzech, K. (b). Upper Napo Kichwa: a documentation of linguistic and cultural practices.
London: SOAS. Http://hdl.handle.net//---C-FFB-A. SOAS,
University of London, Endangered Languages Archive ( December, ).
Grzech, K. (c). Epistemic primacy, Common Ground management and the epistemic
perspective domain. In H. Bergqvist & S. Kittilä (Eds.), Evidentiality, egophoricity and
engagement. Studies in Diversity Linguistics (pp. –). Berlin: Language Science
Press. https://doi.org/CitetononCRdoi:10.5281/zenodo.3975797
Grzech, K. (). Using discourse markers to negotiate epistemic stance: A view from situated
language use. Journal of Pragmatics , –.
https://doi.org/10.1016/j.pragma.2021.02.003
Egophoricity and evidentiality
Hale, A. (). Person markers: Finite conjunct and disjunct verb forms in Newari. In
R. L. Trail (Ed.) Papers in Southeast Asian Linguistics . Pacic Linguistics Series A, .
Canberra: Australian National University, –.
Haspelmath, M. (). Comparative concepts and descriptive categories in cross-linguistic
studies. Language , –. https://doi.org/10.1353/lan.2010.0021
Hayano, K. (). Claiming epistemic primacy: yo-marked assessments in Japanese. In
T. Stivers, L. Mondada & J. Steensig (Eds.). The Morality of Knowledge in Conversation
(pp.–). Cambridge: CUP. https://doi.org/10.1017/CBO9780511921674.004
Heritage, J. (). A change-of-state token and aspects of its sequential placement. In
J.M. Atkinson (Ed.), Structures of Social Action (pp. –). Cambridge: CUP.
https://doi.org/10.1017/CBO9780511665868.020
Heritage, J. (a). The epistemic engine: Sequence organization and territories of knowledge.
Research on Language and Social Interaction .
https://doi.org/10.1080/08351813.2012.646685
Heritage, John. (b). Epistemics in Action: Action Formation and Territories of Knowledge.
Research on Language & Social Interaction (), –.
https://doi.org/10.1080/08351813.2012.646684
Heritage, J. & Raymond, G. (). The Terms of Agreement: Indexing Epistemic Authority
and Subordination in Talk-in-Interaction. Social Psycholog y Quarterly (), –.
https://doi.org/10.1177/019027250506800103
Hill, N.W. & Gawne, L. (). The contribution of Tibetan languages to the study of
evidentiality. In N.W. Hill & L. Gawne (Eds.) Evidential Systems of Tibetan Languages.
Trends in Linguistics. Studies and Monographs (pp. –). Berlin: Mouton de
Gruyter. https://doi.org/10.1515/9783110473742-001
Hintz, D.J. & Hintz, D.M. (). The evidential category of mutual knowledge in Quechua.
Lingua. Essays on Evidentiality –, –. https://doi.org/10.1016/j.lingua.2014.07.014
INEC, Instituto Nacional de Estadística y Censos (). Censo Nacional: Población Y
Economía. Instituto Nacional de Estadística y Censos. http://www.ecuadorencifras.gob.ec
/censo-de-poblacion-y-vivienda/. (Accessed November ).
Janhunen, J. (). Typological interaction in the Qinghai linguistic complex. Studia
Orientalia , –.
Kamio, A. (). Territory of Information. Amsterdam: John Benjamins.
https://doi.org/10.1075/pbns.48
Kendrick, K. H., Brown, P., Dingemanse, M., Floyd, S., Gipper, S., Hayano, K., Hoey, E. et al.
(). Sequence organization: A universal infrastructure for social action. Journal of
Pragmatics , –. https://doi.org/10.1016/j.pragma.2020.06.009
Loughnane, R. (). A Grammar of Oksapmin. PhD Thesis. University of Melbourne.
Michael, L.D. (). Nanti Evidential Practice: Language, Knowledge, and Social Action in
an Amazonian Society. PhD Thesis. University of Texas, Austin.
Michael, L.D. (). Rethinking the communicative functions of evidentiality: Event
responsibility in Nanti (Arawakan) evidential practice. Cadernos de Etnolingu
ı
stica (),
–.
Mithun, M. (). The Languages of Native North America. Cambridge: CUP.
Mushin, I. (). Evidentiality and epistemological stance: narrative retelling. Pragmatics &
Beyond new ser. . Amsterdam, Philadelphia: John Benjamins.
https://doi.org/10.1075/pbns.87
Erika Sandman and Karolina Grzech
Mushin, I. (). Making knowledge visible in discourse: Implications for the study of
linguistic evidentiality. Discourse Studies (), –.
https://doi.org/10.1177/1461445613501447
Mushin, I. (). Editorial: Turn design and epistemic management in small communities.
Journal of Pragmatics , –. https://doi.org/10.1016/j.pragma.2022.03.007
Nikolaeva, I. (). A Grammar of Tundra Nenets. Berlin, Boston: Mouton De Gruyter.
https://doi.org/10.1515/9783110320640
Norclie, E. (). Egophoricity and evidentiality in Guambiano (Nam Trik). In S. Floyd,
E. Norclie & L. San Roque (Eds.), Egophoricity. Typological Studies in Language
(pp. –). Amsterdam: John Benjamins. https://doi.org/10.1075/tsl.118.10nor
Nuckolls, J.B. & Michael, L. D. (Eds.). (). Evidentiality in interaction. John Benjamins
Current Topics volume . Amsterdam, Philadelphia: John Benjamins.
https://doi.org/10.1075/bct.63
Oswalt, R. L. (). The evidential system of Kashaya. In W.L. Chafe & J. Nichols (Eds.),
Evidentiality: The Linguistic Coding of Epistemology (pp. –). Norwood, N.J., Ablex.
Pomerantz, A. (). Agreeing and disagreeing with assessments: some features of
preferred/dispreferred turn shapes. In J. M. Atkinson & J. Heritage (Eds.), Structures of
Social Actions (pp. –). Cambridge: CUP.
Sacks, H. (). In G. Jeerson (Ed.). Lectures on conversation. Vol. . Oxford: Blackwell.
San Roque, L. (). Using you to get to me – Addressee perspective and speaker stance in
Duna evidential marking. Language Typology and Universals, –.
https://doi.org/10.1515/stuf-2015-0010
Sandman, E. (). A Grammar of Wutun. PhD Thesis. Helsinki: University of Helsinki.
Sandman, E. (). Chapter . Egophoricity in Wutun. In S. Floyd, E. Norclie &
L. San Roque (Eds.) Egophoricity. Typological Studies in Language (pp. –).
Amsterdam: John Benjamins. https://doi.org/10.1075/tsl.118.06san
San Roque, L., Floyd, S. & Norclie, E. (). Egophoricity: An introduction. In S. Floyd,
E. Norclie & L. San Roque (Eds.) Egophoricity. Typological Studies in Language (pp.
–). Amsterdam: John Benjamins. https://doi.org/10.1075/tsl.118.01san
San Roque, L. & Schieelin, B. (). Chapter . Learning how to know: Egophoricity and
the grammar of Kaluli (Bosavi, Trans New Guinea), with special reference to child
language. In S. Floyd, E. Norclie & L. San Roque (Eds.), Typological Studies in
Language, vol. (pp. –). Amsterdam: John Benjamins.
https://doi.org/10.1075/tsl.118.14san
Schultze-Berndt, E. (). Shared vs. Primary Epistemic Authority in Jaminjung/Ngaliwurru.
Open Linguistics (), –. https://doi.org/10.1515/opli-2017-0010
Slater, K.W. (). A Grammar of Mangghuer. London, New York: Routledge Curzon.
Slater, K.W. (). Introduction: Language contact in the Amdo Sprachbund. Himalayan
Linguistics, Special Issue (), –. https://doi.org/10.5070/H920355370
Stivers, T., Eneld, N.J., Brown, P., Englert, Ch., Hayashi, M., Heinemann, T., Hoymann, G.,
Rossano, F., de Ruiter, J. P., Yoon, K.-E., Levinson, Stephen C. (). Universality and
cultural specicity in turn-taking in conversation. Proceedings of the National Academy of
Science , –. https://doi.org/10.1073/pnas.0903616106
Stivers, T., Mondada, L. & Steensig, J. (Eds.). (a). The Morality of Knowledge in
Conversation. Cambridge: CUP. https://doi.org/10.1017/CBO9780511921674
Egophoricity and evidentiality
Stivers, T., Mondada, L. & Steensig, J. (b). Knowledge, morality and aliation in social
interaction. In T. Stivers, L. Mondada & J. Steensig (Eds.), The Morality of Knowledge in
Conversation (pp. –). Cambridge: CUP. https://doi.org/10.1017/CBO9780511921674.002
Sun, J.T.-S. (). Evidentials in Amdo Tibetan. Bulletin of the Institute of History and
Philology, Academia Sinica (), –.
Sun, J.T.-S. (). Evidentials and person. In A.Y. Aikhenvald (Ed.), The Oxford Handbook of
Evidentiality (pp. –). Oxford: OUP.
Torero, A. (). Los dialectos quechuas. Anales Cientícos de la Universidad Nacional
Agraria, Lima, –.
Tournadre, N. (). Arguments against the concept of ‘conjunct’/’disjunct’ in Tibetan. In
B. Huber, M. Volkart, P. Widmer & P. Schwieger (Eds.), Chomolangma, Demawend und
Kasbek: Festscri für Roland Bielmeier zu seinem . Geburtstag (pp. –). Bonn:
VGH Wissenschasverlag.
Tournadre, N. & LaPolla, R. J. (). Towards a new approach to evidentiality: issues and
directions for research. Linguistics of the Tibeto-Burman Area (), –.
https://doi.org/10.1075/ltba.37.2.04tou
Widmer, M. (). The evolution of egophoricity and evidentiality in the Himalayas: The case
of Bunan. Journal of Historical Linguistics (), –.
https://doi.org/10.1075/jhl.7.1-2.10wid
Willett, T. (). A Cross-Linguistic Survey of the Grammaticization of Evidentiality. Studies
in Language (), –. https://doi.org/10.1075/sl.12.1.04wil
Zeisler, B. (). Don’t believe in a paradigm that you haven’t manipulated yourself! –
Evidentiality, speaker attitude, and admirativity in Ladakhi. Himalayan Linguistics (),
–. https://doi.org/10.5070/H917136797
Address for correspondence
Erika Sandman
University of Helsinki
Department of Languages
P.O. Box (Unioninkatu )
Helsinki
Finland
erika.sandman@helsinki.
Co-author information
Karolina Grzech
University of Valencia
Faculty of Philology, Language and
Communication
Department of Spanish Philology
karolina.grzech@uv.es
Erika Sandman and Karolina Grzech
Publication history
Date received: November
Date accepted: April
Published online: August
Egophoricity and evidentiality
