Egophoricity in Central Stau copular clauses

Abstract

This article demonstrates that the Central Stau language (Horpa < Gyalrongic < Sino-Tibetan/Trans-Himalayan) possesses a binary egophoricity contrast in its copular system in affirmative clauses not described and analysed in detail before. It examines the functions of the egophoric copula ŋu and its non-egophoric counterpart ŋə. Of these, the former signals the relevant speech act participant’s personal involvement, epistemic authority or a portrayed stance of a close bond in the proposition. Despite the differences in their functions and differing prototypical domains of use, reflecting patterns of “canonical” egophoricity, the choice between the copulas shows great flexibility and frequently reflects how the speakers wish to encode their epistemic stance. In brief, situation-dependent discourse pragmatics, rather than grammatical person encoded by the copular subject, determines copular use in Stau. The article thus concurs with other recent research on egophoricity that highlights the versatility of this epistemic category.

1. Introduction

This article investigates copular clauses in the Central Stau language (Horpa < West Gyalrongic < Gyalrongic < Sino-Tibetan/Trans-Himalayan) (henceforth Stau).¹ With the help of primary source materials, and a focus on epistemicity, it argues that the copular clauses manifest an egophoric contrast not discussed and analysed in these terms and to a sufficient extent in earlier research. The introduction defines the goals of the article (1.1), briefly discusses its sources and methodology (1.2) and concludes by laying out the structure of the article (1.3).

1.1. Goals of the article

This fieldwork-based article offers an analysis of egophoricity in Stau copular clauses. It builds on earlier and currently rapidly growing West Gyalrongic scholarship, such as Gates (2021) on Central Stau, Honkasalo (2019) on Eastern Geshiza and Tunzhi (2019) on Zhangda Northwestern Stau copular systems.

The article has the following three primary goals. First, in order to lay the foundations for later analysis and the arguments, it identifies the copular forms and their argument indexation properties in Stau. The findings are compared with the interpretations of earlier research. Second, by analysing the use of the Stau copulas, the article demonstrates that the distribution of the affirmative copulas exhibits an intersubjective phenomenon analysable with the notion of egophoricity. Finally, with accompanying illustrative examples, it describes the use and functional domains of both the egophoric and non-egophoric² copulas in Stau. It also compares the system of Stau with “canonical” egophoricity established in the first wave of egophoricity research and with the principles of “flexible” egophoricity gaining focus in more recent scholarship.

Egophoricity has been researched intensively in Sino-Tibetan in recent decades, although the grammatical notion is now receiving increasing attention outside the language family as well. Besides the well-known cases of the Tibetic languages with almost universal egophoricity (see Tournadre and Suzuki 2023: 416), egophoricity is attested in various other branches of the language family, such as Himalayish (e.g. Newar, see Hale 1980), “Qiangic” (e.g. Wadu Pumi, see Daudey 2014) and Sinitic (e.g. Wutun, see Sandman 2016).

Egophoricity in West Gyalrongic, however, remains a neglected topic, both in terms of the copular systems and more generally in grammatical investigations of the relevant languages. Also, the simultaneous presence of argument indexation and egophoricity is typologically very rare, and the Gyalrongic branch of Sino-Tibetan constitutes the only case with fully fledged person indexation and egophoricity coexisting in the language family (Jacques 2019: 515–16). As argued in the article, this applies to Stau as well. While earlier studies report no egophoricity in the language, its relatively complex argument indexation system has been described in detail by Gates (2021) and Honkasalo (2020). In sum, this article addresses the existing lacuna in current research and, at the same time, aims to expand our knowledge about egophoricity and copular behaviour in Horpa languages. As the brief comparison towards the end of the article suggests, egophoricity may be present more broadly among the Horpa languages.

1.2. Sources and methodology

In addition to the author’s fieldwork in the Stau homeland, most of the source materials on which the article builds have been collected ex situ in India where Stau exile communities reside. All research partners in India have left the Stau homeland (Ch. 道孚县, Tib. rta’u rdzong) as adult native speakers and continue to use the language daily. It goes without saying that while working with exile communities is not free from risk, the form of Stau that appears in this article is both grammatically and lexically highly homogeneous with other existing descriptions of the language, including Gates (2021). In addition to staying in communities with other Stau speakers whenever this is feasible, the exiled Stau frequently take advantage of the possibilities offered by smartphone applications, such as the Chinese WeChat (微信), to stay in touch with other Stau speakers. Such activities clearly support language maintenance among Stau migrants. The situation, however, differs strikingly among the ethnic Stau born abroad, among whom the intergenerational transmission of Stau is now at risk.

The principal Stau partners in this research project originate from Poxiu (Ch. 坡修, Tib. phog sho) and Yepo (Ch. 也坡, Tib. gyas phyogs) villages of Mazi (Ch. 麻孜, Tib. ma zur) in the Stau homeland. To indicate their respective contributions in the examples, the former is coded as (P) and the latter as (Y). The examples offered in this article are a fortuitous side-product of the author’s morphosyntactic and lexicographic study of the Stau language,³ rather than materials collected with the specific aim of copular clause research.

In collecting the source materials, in addition to Stau, the author has used Dharamshala Central Tibetan as a lingua franca. The fact that Central Tibetan possesses egophoricity makes it convenient to explain certain features of the Stau copular system through this auxiliary language. On rare occasions, Y also expressed some of his insights in English. Due to its well-known risks, the article does not apply translation elicitation whereby the research partners would be asked to translate sentences from any language into Stau as the main method of investigation. Translation from Tibetan to Stau, however, has been used to further clarify and illustrate contexts where a phenomenon has been identified from primary data. In addition, acceptability judgements have been frequently implemented and indicated as such in the examples. As the article argues, egophoricity in Stau is largely a discourse-pragmatic phenomenon, and the acceptability judgements should be seen from this perspective.

1.3. Structure of the article

This article has the following structure. Section 2 addresses the theoretical key concepts the article relies on, namely egophoricity and territory of information. Then, Section 3 offers an overview of the copular system in Stau and discusses earlier studies of copulas in the language, highlighting existing issues that the article aims to tackle. In turn, Section 4, which forms the core of the article, investigates the use of the two Stau copulas and reflects the findings against “canonical” egophoricity and discourse-pragmatic based stance taking. Section 5 compares the copular system of Stau with those of other Gyalrongic languages, arguing for the need to pay attention to egophoricity in future research. Finally, Section 6 concludes the article by summarizing its key findings and arguments.

2. Key concepts: egophoricity and territory of information

This section is dedicated to the two key concepts this article relies on: territory of information (2.1) and egophoricity (2.2).

2.1. Territory of information

The article follows the pioneering study of the Japanese linguist Akio Kamio concerning “territory of information” as an epistemic and interpersonal notion. As defined by Kamio (1997: 2), information that speakers consider “proximal” to themselves falls into their territory of information. For instance, it is natural for a speaker to say that she is 36 years old and lives with her two children in Daofu, the Stau homeland. In many languages, however, such a statement requires a highly specific pragmatic condition to be uttered in the second person, a potentially unacceptable intrusion into the epistemic territory in which one has no authority. Analogous to territorial animals (or pre-modern states and other political entities), participants in a linguistic discourse constantly track their own and other participants’ territories of information that are continuously negotiated. The “heartland” in territories of information, such as a speaker’s knowledge regarding his name, identity and other matters to which he has primary epistemic access and high epistemic authority, remain uncontested most of the time. On the other hand, more peripheral territory can be gained, lost, shared or even trigger a “territorial conflict” (Honkasalo 2023: 206; see also Bristol and Rossano 2020 on the hierarchical structure of epistemic territory). In all, as a metaphor, the notion of “territory of information” aids in conceptualizing matters of epistemic authority and primacy relevant for this article.

2.2. Egophoricity

Also known as conjunct/disjunct marking in earlier research, “egophoricity” in broad terms refers to the linguistic encoding of “personal knowledge, experience, or involvement of a conscious self” (San Roque et al. 2018: 2).⁴ Phenomena with some or all of these characteristics are sometimes discussed under the label of evidentiality (i.e. grammaticalized encoding of the information source; see Aikhenvald 2004), and the mutual relationship of evidentiality and egophoricity remains debated. The article considers evidentiality and egophoricity as separate grammatical categories that nevertheless possess similar discourse functions, a view adopted from Sandman and Grzech’s (2022) insightful comparison of Wutun (divergent Sinitic) and Upper Napo Kichwa (Quechuan).⁵ Similarly, egophoricity should be regarded as distinct from person marking, despite some functional overlap between the two (Bergqvist and Kittilä 2017, 2020: 3).

While the phenomenon is also attested in other language families, Sino-Tibetan languages form the context in which egophoricity has been studied most extensively. A pair of examples from Central Tibetan below briefly illustrates the key principle of “canonical” egophoricity. In (1), the speaker uses the egophoric copula yin, since he discusses his self-identity, and thus has higher epistemic authority, privileged access to knowledge and personal involvement in the matter expressed by the proposition. On the other hand, in (2), the speaker uses the non-egophoric form red, since the identity of a third person outside the EGO of the speaker is discussed and the four conditions listed for (1) above are not fulfilled, at least not in full and to the same extent.

3. Stau copulas and previous research

A copular clause includes a copular subject (CS) and a copular complement (CC) that, in addition to a “zero copula”, may be linked with a copula, the formal properties of which vary across languages of the world. Stau is a language with compulsory copular use where the copulas manifest the morphosyntactic properties of verbs. The Stau copular system includes three copulas: the affirmative copulas ŋə and ŋu together with the suppletive⁶ negative copula mɲæ ∼ mjæ, as identified by Gates (2021: 379) in his descriptive grammar of Mazi Stau.⁷ Wengmu (2019: 173) also addresses the copular system of Stau by identifying ŋə as an affirmative and mja as a negative copula in the language, although her grammatical treatise only briefly deals with the copulas. In addition, Wengmu’s examples include a form Gates and this article classify as copula, namely ŋu (3), but no further explanation regarding this form is offered. In sum, prior to this article, Gates’s (2021: 379–84) analysis of Mazi Stau copulas presents the most extensive analysis of the copular system and copular clauses in Stau.

In Gates’s (2021: 379) analysis, ŋu is termed as an SAP copula, indexing “first and second person singular with ŋu, first person plural with ŋõ ⁸ and second person plural with ŋun”. For instance, in (4), the first person pronoun is SAP, which consequently triggers the use of the copula ŋu. On the other hand, in (5), the proposition falls outside the scope of SAPs, with the result that the non-SAP copula ŋə must be used.

In sum, the SAP copula is restricted to speech-act participants, namely, first and second person.⁹ While the singular forms are identical, a contrast arises in the plural forms. Gates (2021: 261) does not illustrate with paradigmatic examples the argument indexation properties of the other two analysed copulas, namely ŋə and mjæ, although he states that copulas belong to the verb class 2b (discussed below) in Stau, a notation already sufficient for grasping how the copulas conjugate. To help the reader, the full paradigm of the interpretation for ŋu is offered in Table 1.

Table 1. Gates’s (2021) analysis of the argument indexation properties of the SAP copula ŋu

On the one hand, Gates’s analysis of ŋu with a focus on SAPs gets close to the egophoric interpretation of this article without applying the term and qualifies for a foundational insight in Stau copular studies. Markers in egophoric systems interact with the speech-act roles (Sandman and Grzech 2022: 82; Bergqvist and Kittilä 2017: 19; see also Dahl 2000 on a broader and slightly different approach to egophoricity as a discourse phenomenon). On the other hand, the analysis results in an anomalous and irregular intransitive verb with number contrasts in addition to person contrasts not attested elsewhere in Central Horpa.¹⁰

Three primary explanations may be offered. First, the Stau language is known for internal variation, and Gates’s sources, which differ from those of this author, might explain the copula’s behaviour.¹¹ Second, copulas often behave in idiosyncratic ways, which may explain the peculiar argument indexation properties claimed for ŋu. Discussing copulas from a typological perspective, Dixon (2010: 178) states that they frequently manifest irregular behaviour. Finally, as this article proposes, it is possible to reanalyse the copular system in an alternative way. As an unintended side product of this analysis, the copulas also appear as completely regular verbs fully compatible with the Horpa verb classification system proposed by Honkasalo (2020).

In addition to the SAP forms, Gates (2021: 379) shows that SAP copula ŋu also appears with third person copular subjects, as in (6). This apparent distributional discrepancy is explained in terms of the SAPs where “I” as the speaker is part of the third person’s company. The copular subject, however, falls outside the scope of SAPs in terms of strictly grammatical encoding. Similarly to Gates’s realization, this article will demonstrate that discourse pragmatic factors, rather than independent morphosyntactic rules, ultimately determine the use of the two copulas in many instances where variation is possible. Finally, in Gates’s terminology, if the non-SAP copula ŋə were used in (6) it would imply that “I” as the speaker is not part of the third person’s company.

This article offers an alternative analysis of the copular system from the viewpoint of egophoricity and territory of information. Stau has three copulas: affirmative egophoric, affirmative non-egophoric and their negative counterpart with no egophoricity distinctions. Table 2 summarizes the existing copular forms and their argument indexation properties, as present in the source materials of this article. As the table demonstrates, egophoricity in Stau exists as a binary system in affirmative clauses, egophoricity distinction being neutralized in negative clauses. This manifests a typological tendency where fewer grammatical distinctions appear in negative than in affirmative clauses (see Miestamo and van der Auvera 2011 for the neutralization of grammatical distinctions in the negative).

Table 2. The forms of the copulas in Stau

Using Honkasalo’s (2020) classification of Horpa verbs, the affirmative egophoric copula ŋu lacks any argument indexation properties and always surfaces in the shape of ŋu, regardless of the person and number of the copular subject, as in (7–11). Consequently, it belongs to the intransitive verb class 1b with other non-stative verbs lacking any argument indexation, such as də “inanimate existential verb” and sko “to manage, be able” (see Honkasalo 2020).

On the other hand, the affirmative non-egophoric copula ŋə belongs to the intransitive verb class 2b, which means that it indexes the person of the copular subject but not its number, like verbs, such as dʑi “animate existential verb” and ɕə “to go”. This results in a paradigm of three distinct forms (12–14).

The negative copula mɲæ with no egophoricity distinctions also belongs to the verb class 2b. As Table 2 demonstrates, and as expected of class 2b verbs, the Stau varieties analysed in this article generally distinguish between first and non-first (namely, second and third) person in intransitive scenarios, which also applies to the negative copula. From this, it follows that most intransitive verbs lack a distinction between second and third person in their argument indexation properties. As argued by Honkasalo (2020) addressing the reduction of paradigmatic forms, this results from a partial loss of argument indexation marking in second person forms historically encoded with the eroding -n, which often renders the second and third person forms identical in intransitive verbs. The second person marking is retained most faithfully with the imperative and prohibitive forms (Honkasalo 2020), as illustrated by (15) where -n indexes the second person.

The negative copula, however, cannot be used in commands (16). As a result, no contexts exist where the second person argument indexation morpheme -n attaches to it, which results in slight formal asymmetry when seen against the copula ŋə with three distinct forms in its paradigm.

4. Use of the copulas

This section forms the core of this article; it describes and analyses the use of the Stau copulas. It is divided into the following sections: egophoricity and person (4.1), egophoricity and the sensory evidential -rə (4.2), egophoricity and the knowable attributes of the first person (4.3), egophoricity and personal relations (4.4), egophoricity with objects and entities (4.5) and egophoricity and general facts (4.6).

4.1. Egophoricity and person

Stau speakers use the egophoric copula ŋu when they wish to express personal involvement or privileged epistemic authority, also known as epistemic primacy (“the right to know and to discuss something”; see Stivers et al. 2011: 13), regarding the proposition, the information of which lies in the territory of information of the SAP whose perspective is adopted. In practice, this is often the first person, the primary domain in all models of “canonical” egophoricity. To illustrate, stating one’s personal identity qualifies as something about which the speaker him or herself has the highest epistemic authority (17).

The non-egophoric copula also appears with the first person. When no evidentials are present, the difference between the two affirmative copulas in this context remains minute, and in many circumstances, the two are interchangeable. The major difference lies in the projected epistemic stance, namely what kind of attitude the speaker expresses regarding the information of the proposition. It is proposed that while (17) emphasizes that being a lama is closely connected to the speaker’s EGO and the speaker thus highlights his personal involvement, the use of the non-egophoric copula presents the same information in a more matter-of-fact fashion (18).¹⁴

As discussed earlier, the egophoric distinctions are neutralized in negative clauses. Both copular clauses are thus negated identically with the negative copula in its dedicated first person form (19).

The mind of the second person remains inaccessible to cognitive outsiders. Consequently, it is more common to ask questions than to make statements when addressing the second person in the Horpa languages. Stau second-person questions manifest a pattern of “canonical” egophoricity termed the “anticipation rule” by Tournadre and LaPolla (2014: 244),¹⁵ also known as “perspective shift” (San Roque et al. 2018). If the topic of interrogation is perceived to lie firmly inside the territory of information of the addressee’s EGO, the egophoric copula is chosen in anticipation of the addressee’s epistemic perspective, as in (20, 21). In the affirmative, the addressee will most likely equally answer with the egophoric copula, as illustrated above in (17).

The negative copula is not compatible with the interrogative prefix æ- (22), resulting in structural asymmetry in negative questions. Instead, infrequently occurring negative questions must be formed with the interrogative enclitic =a, as in (23).

While it was stated above that questions, rather than statements, are more common in the second person, statements also occur. This is possible in circumstances where the information is not exclusively in the epistemic territory of the addressee, such as describing the addressee with adjective copular complements. In (24), the speaker states a value judgement concerning the addressee and uses the non-egophoric copula. In contrast, (25) does the same with the egophoric copula. Again, the pragmatic effect of the two differs and is explainable by means of projecting a different epistemic stance, and (24) takes the form of a general statement paying no attention to the interpersonal relationship between the speaker and the addressee. Conversely, (25) is deemed pragmatically ungrammatical unless the addressee has a particularly strong relationship with the speaker, such as being a family member of a close friend.¹⁷ In sum, the speaker has epistemic authority to judge the addressee’s character by using the egophoric form only in cases of strong personal involvement and personal bonds.

Both the egophoric and non-egophoric forms occur with the third person as the copular subject, which proves that the contrast between the copulas in Stau should not be interpreted from the viewpoint of person marking/argument indexation of the grammatically encoded SAPs.¹⁸ In (26), the speaker knows that there is no money in his pocket. If, on the other hand, this comes as a sudden realization, for instance by putting one’s hand inside the pocket to take out the money that should be there but failing to do so, the egophoric form is deemed ungrammatical. Instead, the non-egophoric copula must be used with the sensory evidential suffix (discussed below), as in (27).

4.2. Egophoricity and the sensory evidential -rə

Stau exhibits both grammaticalized egophoricity and evidentiality, which makes the language important for the study of these categories. Moreover, the two interact with each other.¹⁹ Before analysing the interplay between the copulas and the sensory evidential, it is necessary to offer an overview of the latter in Stau. The sensory evidential is used for marking information that prototypically lies outside the EGO and is gained through the senses, as in (28) where it indicates that the speaker saw the flying bat.²⁰

In addition, the evidential possesses a special use with information directly in the domain of the EGO where it encodes “endophatic processes”, a term coined by Tournadre (1996a: 226; 1996b: 206–07) to refer to internal sensations that cannot be observed directly by outsiders, such as such as hunger (29), thirst (30), pain (31) and fear (32).

Finally, the sensory evidential accompanies states and actions, over which the subject has no direct control, such as the mental activities of liking and understanding something, as in (33) and (34) (cf. also Honkasalo 2019: 603 on Geshiza).

Returning to the interplay of the copulas and evidentiality, the use of the sensory evidential suffix -rə encoding information received through the senses renders the statements (17–19) ungrammatical (35–37). Personal identity belongs to the core of one’s territory of information, yet the sensory evidential typically encodes information that one has received externally through the senses. As a result, the egophoric copula ŋu is incompatible with the sensory evidential. While it is also pragmatically ungrammatical with the non-egophoric form in (36), the article will subsequently demonstrate that contexts exist where the two are compatible.

Table 3 summarizes the compatibility of the copulas with the sensory evidential suffix -rə. As can be seen, all restrictions concern the affirmative egophoric copula only, since both the affirmative non-egophoric copula ŋə and the negative copula mɲæ may host the evidential suffix -rə in all their personal forms.

Table 3. Compatibility of the copular forms with the sensory evidential suffix -rə

To illustrate briefly, in (38), a mother comments on her child’s tendency to bite. Rather than emphasizing the close connection between the mother and the child, discussed in the next section, the focus is found in the mother’s sensory observation of her child’s behaviour expressed by the suffix -rə. In turn, in (39), the speaker expresses a generally known fact that cuckoos start appearing in Daofu in April and marks this piece of information with the sensory evidential suffix that here plays a broader role of encoding general knowledge, discussed at the end of this section.

To conclude, Gates (2021: 383) states that the sensory evidential obligatorily accompanies the copula ŋə in non-SAP copula verb constructions, which roughly correspond to the non-egophoric copular clauses of this article (see Section 3).²¹ Our investigation confirms this distributional insight. At the very least, it is a very strong tendency. Exclamations where the modal discourse enclitic =væ can be thought to replace the sensory evidential constitute the only potentially unambiguous exception in the source materials, as in (40), but it remains to be analysed further whether such exclamations can be classified as canonical copular clauses.

In any case, the frequent co-occurrence of the non-egophoric copula and the sensory evidential in the third person non-SAP forms arises naturally from the fact that since the speaker admits lacking direct epistemic access or personal involvement by not using the egophoric copula, an evidential is needed to specify the source of information.

4.3. Egophoricity and the knowable attributes of the first person

Sometimes personal information exists in the territory of information of the EGO, but it is still relatively accessible to the addressee, which downplays the EGO’s privileged epistemic access. Thus, such information cannot be considered to be located at the core of the territory of the speaker, but more at the periphery, with potentially some territorial overlap with the addressee. In such cases, the Stau copular complement is typically an adjective characterizing a property of the copular subject. To illustrate, while one’s stature or body weight are closely connected with one’s EGO, they are also something observable to outsiders. Consequently, the speaker may adopt a non-EGO perspective to describe his or her stature (41), but equally, the egophoric copula may be used in this context (42), depending on one’s stance whether to underline a strong connection with the EGO or not.

In (43), a speaking trickster rabbit in a well-known folktale warns the boy protagonist that he (the rabbit) is very heavy. The utterance is encoded with the non-egophoric form of the copula. As it happens, the same story has been recorded by Gates (2021) where the egophoric copula appears at the same point in the folktale (44). In sum, the property of weight is something that can often be observed outwardly, at least to an extent, which justifies the alternation attested here.

4.4. Egophoricity and personal relations

When the bond between people is represented as very strong, the use of the egophoric copula ŋu can be extended to other people outside the EGO, encoded as third person forms grammatically. This use, going beyond “canonical” egophoricity, corresponds to Gawne’s (2017) “broad scope” of egophoricity.

The “broad scope” use occurs particularly frequently with one’s family members in Stau. In (45), the speaker discusses her maternal aunt who is skilled at knitting sweaters. Although the maternal aunt is clearly distinct from the speaking SAP and a separate person, the use of the egophoric form emphasizes the closeness of the relationship between the two, which grants the speaker epistemic authority to discuss her aunt with the egophoric coding of knowledge.

There are limits to the usability of the egophoric forms when discussing people, related to the notions of epistemic upgrading and downgrading. Rather than being binary (i.e. either/or), epistemic access may manifest to varying degrees, and speakers may thus “upgrade” or “downgrade” the access they claim, e.g. by tag questions and uncertainty markers (maybe, probably) in English (Stivers et al. 2011: 12).

In (46), the husband of the addressee is praised for his skills as a carpenter. Since the husband and his skills clearly fall within the territory of information of the addressee, rather than the speaker, the non-egophoric copula (46) and not the egophoric copula (47) must be used so that the speaker does not intrude into the epistemic territory of the addressee and claim it as her own. In other words, the speaker lacks epistemic authority to state (47), which renders it pragmatically incorrect in most circumstances.²³ In sum, the speaker “downgrades” her own authority by choosing the non-egophoric copula.

When discussing personal relations, the choice of the copula serves as a mechanism of stance taking, for which I borrow Häsler’s (1999: 151–52) terms “strong empathy” and “weak empathy”.²⁴ Both (48) and (49) below convey the moral judgement that the speaker’s friend is a bad person. The choice of the copula, however, conveys additional information concerning the speaker’s relationship with the addressee. With the egophoric form, the speaker wishes to express that the relationship between the two is strong, a manifestation of strong empathy. Despite his or her flaws, the person remains the speaker’s dear good friend. On the other hand, adopting the non-egophoric copular form indicates that the friend is not particularly close to the speaker, a manifestation of weak empathy.

The example pair (50) and (51) offers a similar scenario, but this time concerning the territory of information of the second person addressee. With (50), the speaker takes the addressee’s viewpoint by anticipating the answer with the egophoric copula. In contrast, the same question can be posed with the use of the non-egophoric copula that does not anticipate the addressee’s viewpoint.

The Stau language has the means to simultaneously mark a piece of information as inside the speaker’s privileged territory of information and yet downplay the closeness of the personal relationship involved. In (52), by using the egophoric copula, the speaker expresses that Akhu Stonpa is a great trickster. Since he, however, is a legendary folklore figure, attempts to use only the egophoric form are deemed as incorrect (53). In such a case, Stau speakers interpret the sentence as incorrect due to its pragmatic implausibility. The speaker claims direct and personal knowledge of Akhu Stonpa who is a legendary folk hero and not a real person who is currently alive.²⁵

To gain pragmatic acceptance among Stau speakers, (52) requires the addition of the suffix -gə to the predicate verb. The suffix plays a modal function of distancing and uncertainty in the utterance. While by using the egophoric copula the speaker claims epistemic authority and deep personal knowledge of Akhu Stonpa, most likely through having heard many folk stories, the suffix -gə indicates that the person has not directly met Akhu Stonpa and thus cannot be quite sure of matters concerning him. The phenomenon can be further illustrated by two replies to a common Stau question (54). In the first reply without -gə, the speaker implies that she has met her mother and is thus sure that she is doing well. On the other hand, the reply with the suffix -gə insinuates that the person supposes that her mother is doing well, but she cannot be fully sure, since she has not met her for some time, which impedes full certainty.

4.5. Egophoricity with objects and entities

As has become increasingly clear, while the egophoric forms typically occur in the context of the first person, they also surface with the second and third persons because of stance projection. This goes beyond human relationships and applies equally to objects and other more abstract entities. To illustrate, the speaker discusses her shoes with the egophoric form in (57), since this underlines the connection between the object and its owner. The shoes were most likely bought by her and, in any case, she is the one who wears them and has deep personal knowledge about them. Consequently, the use of the egophoric copula instead of the non-egophoric copula emphasizes the bond between the object and its owner.

In (58), the speaker originates from Daofu, and can thus be expected to know her home thoroughly. Consequently, the statement is coded with the egophoric copular form. On the other hand, the non-egophoric counterpart of this sentence in (59) can be uttered by an outsider, such as the author, who is not a native Daofu, but nevertheless possesses some limited knowledge of the place. The use of the non-egophoric form demonstrates to the addressee the speaker’s weaker connection with the place. Again, the choice of the copula enables the speakers to express the degree of their epistemic authority and bond with the information offered in the proposition.

In (60), the speaker takes the addressee’s perspective and asks about her village with the egophoric form. This is because, in the case of an (affirmative) answer, the egophoric copula is expected to be used.

The example pair (61) and (62) provides a further point of comparison. Both utterances convey the meaning that eggplant dishes are somewhat tasty. In (61), however, the speaker has cooked eggplant dishes before so that her familiarity with the dish justifies coding the utterance with the egophoric form. This highlights the speaker’s involvement in the cooking process, resulting in privileged knowledge and authority to discuss the taste, unlike someone who cooks such a dish for the first time. Using the egophoric form as a first-time cook would mislead the addressee due to its implications of earlier experience.

In contrast, (62) represents eggplant dishes’ tastiness in a general fashion with no implied involvement of the speaker or her earlier experience. This relates directly to the next topic of the article, namely the expression of general facts, discussed below.

4.6. General facts

Following Kittilä (2019), general knowledge refers to information that belongs to the speaker’s established world view, is referable without any external evidence and its original sources do not need to or even cannot be specified.²⁶ Stau encodes general knowledge using the sensory evidential discussed earlier in this article.

The egophoric copula cannot be used for general facts that are supposed to be known for everyone and thus cannot be located inside any individual speaker’s private territory of information, but rather in an epistemic shared territory everyone can access. To illustrate, (63) states a general fact, namely that rabbits are fast-running animals, something most people with at least cursory knowledge regarding them can reasonably be expected to know. Consequently, when referring to rabbits as a class of animals, the use of the egophoric copula is deemed ungrammatical by Stau speakers (64). An attempt to do so is tantamount to claiming that the information regarding rabbits’ fast running belongs to the speaker’s territory of information only. Further examples regarding the interplay of the sensory evidential and the non-egophoric copula in marking general knowledge are given in (65) and (66).

The use of the copula also affects the interpretation of the utterance. In (67), the non-egophoric copula results in a general interpretation applicable to fields in general: “Fields are very green.” Replacing the copula with its egophoric counterpart requires the interpretation that the speaker is discussing his/her own fields (68). This is because the other interpretation, namely that the speaker makes a claim that the greenness of fields is something exclusive to his/her territory of information, is pragmatically not expected and thus considered incorrect, as in (64) above that can only be considered pragmatically correct in a peculiar case where the utterance refers to a particular rabbit as an individual and a strong bond of some kind exists between it and the speaker. Perhaps the speaker has a pet rabbit s/he loves dearly, resulting in a special bond between the two. This would render (64) readable as “(My dear pet) rabbit is fast.”

Moreover, in (69), the speaker uses the non-egophoric copula to make a claim concerning a general fact supposedly known to speakers of Stau who live in an environment where Chinese alcoholic drinks are widely available. In contrast, the use of the egophoric copular form emphasizes that the speaker knows from his personal experience that Chinese baijiu (白酒) alcohol is strong (70). In other words, the statement ceases to be merely a general statement and becomes more personal.

5. Comparative remarks

This section briefly compares Stau with other Gyalrongic languages in terms of their copular systems. It argues that while this article focuses on Central Stau, egophoricity distinctions can possibly be identified in other related Gyalrongic languages as well, which highlights the need for further research.

Table 4 provides a comparison of affirmative and negative copulas in selected Gyalrongic languages.²⁷ The listing is not meant to be exhaustive, and since several of the languages remain seriously under-researched, future investigation may consequently discover new copulas or attribute new functions to already known copulas. It is not possible illustrate each language in detail in this article. In what follows, I offer some highlights and underline the possibility for discovering Central Stau like phenomena of egophoric contrast.

Table 4. A comparison of copulas in Gyalrongic languages

Honkasalo (2019: 476–79, 650–53) discusses copular clauses in Geshiza. The language possesses two copulas: the affirmative ŋuə (71) and its negative counterpart mɲa (72). Unlike Stau, however, the language lacks egophoricity distinctions in its copular system. The same seems to apply to Dandong (personal fieldwork), although for a full confirmation, a thorough morphosyntactic description of this understudied Horpa variety is needed. Dandong possesses some Stau-like features (see also Gates et al. 2022: 214 on the geographically close Dangling variety as a “hinge dialect” between Geshiza and Stau), but in its copular system, the lect resembles Geshiza more.

The Geshiza general affirmative copula ŋuə is a direct cognate with the Stau non-egophoric copula ŋuə. Their argument indexation properties are almost identical, as Table 5 illustrates (see Table 2 for comparison). In Geshiza, the second person form mɲa-n nevertheless retains the morpheme -n discussed earlier in this article.

Table 5. Argument indexation properties of the Geshiza copulas

Tunzhi (2019: 292–99) identifies two affirmative copulas for Northwestern Stau: ŋo and to, and based on the available examples, ^mɲa, a cognate with the Central Stau negative copula, can be analysed to share a copular function. In Tunzhi’s interpretation, ŋo is a suppletive form of to in the first person. As the offered examples illustrate, however, both copulas can be used with the first person, highlighting their flexibility. The roles and functions of the two copulas appear strikingly similar to egophoric distinctions in Stau analysed in this article. In other words, to resembles a non-egophoric copula that codes the act of giving neutrally in (73) while ŋo manifests an egophoric function by highlighting the close connection with the speaker’s EGO in the act of giving (74). Northwestern Stau remains under-researched, and the issue of egophoricity needs to be addressed in detail in future studies.

Together with Horpa, the Khroskyabs languages form the second branch of West Gyalrongic. They resemble Horpa in copular use. For instance, Wobzi Khroskyabs possesses a binary system with an affirmative copula and its negative counterpart (Lai 2017: 247–49). No egophoricity distinctions, however, have been identified.

Equally, East Gyalrongic languages possess multiple affirmative copulas (see Jacques 2021 on ŋu and ɕti), yet they seem not to make egophoric distinctions in their copular system. In sum, the Horpa languages, particularly those with a close relationship with Central Stau, such as Northwestern Stau and Erkai, offer the “lowest hanging fruit” for ascertaining the spread of egophoricity in the language cluster. In addition, other languages of the Horpa cluster should be investigated concerning this feature. Also, further investigation of the Khroskyabs varieties is necessary to analyse conclusively the spread of egophoricity in West Gyalrongic.

6. Conclusion

Building on earlier copula studies in the Horpa lects, particularly Gates (2021) on Central Stau, Honkasalo (2019) on Geshiza and Tunzhi (2019) on Northwestern Stau, the present fieldwork-based article offers the first comprehensive description of Central Stau copulas from an interpersonal epistemic viewpoint. Through an analysis of the Stau copular system, the article argued that the language manifests distinctions in copular choice commonly discussed in terms of egophoricity. Although it showed the connection of egophoricity with SAPs, co-occurrence with the SAPs is not a defining feature of the copula ŋu. Consequently, Gates’s term “SAP copula” for ŋu can be adjusted as “egophoric copula”, which yields an additional benefit of making the discussed phenomenon more recognizable and comparable typologically.

Studies on egophoricity have proposed a “canonical” egophoric patterning whereby the first person statements and the second person questions as a result of an “anticipation rule” accompany an egophoric form, a non-egophoric form being used elsewhere. As San Roque et al. (2017) explain, this perspective shift forms the focus in “canonical” definitions of egophoricity. Table 6 summarizes this distribution and illustrates how it is realized in Stau. As this table and the examples in this article demonstrate, Stau copulas can appear in a distributive pattern expected in “canonical egophoricity” and often do so.

Table 6. “Canonical” egophoric patterning and the Stau affirmative copular forms

Nevertheless, Table 6 by itself would offer a distorted idea of copular distribution in Stau, forcing it into a preconceived idea regarding how egophoric distribution “must” look in a language. This article revealed that, in many instances, both the egophoric and non-egophoric copular forms are acceptable, and their distribution manifests discourse-pragmatically controlled flexibility of a high degree.

A comparison of Table 7 with Table 6 underlines this considerable flexibility. In other words, the speakers often have the freedom to choose between the egophoric and non-egophoric forms, depending on how they wish to encode their epistemic stance towards the proposition. In this, the findings of this article align with those of other recent research on egophoricity, such as Sandman and Grzech (2022). Despite the frequent egophoric marking in the contexts of first person affirmative statements and second person questions as a result of a “perfective shift”, egophoricity in Stau is independent from grammatical person and may be used with any person as the copular subject.³¹ To conclude, the ultimate determining factor in copular distribution in the language lies in the speaker’s desire to signal his or her epistemic stance vis-à-vis the proposition. Accordingly, when applying the parameter of “flexibility” discussed by San Roque et al. (2018: 27–28), Stau clearly qualifies as a language with flexible, rather than rigid, egophoric marking.³²

Table 7. Additional distribution of the affirmative copular forms in Stau

While copular egophoricity exists in other Qiangic languages and more broadly in other Sino-Tibetan languages, Stau offers the first documented instance of the phenomenon in the Horpa language cluster. Hale’s (1980) research on Kathmandu Newar has been pivotal in shaping the image of how egophoricity is conceptualized, but as subsequent investigations of egophoric systems have demonstrated, egophoricity manifests variation in the purportedly defining features proposed for the grammatical category in earlier research (Bergqvist and Knuchel 2017). Also, egophoric copulas have been investigated most extensively in Tibetic languages. Against this backdrop, Stau from a different branch of Sino-Tibetan and with an evolved argument indexation system often absent in egophoric languages offers not only new perspectives to the research on Sino-Tibetan egophoricity but also contributes to the ongoing task of clarifying the key features and functional domain(s) of egophoricity.