|Home||Linguistik online 36, 4/2008|
Distance and Visibility: Two Systems in Hausa Deixis *
Mahamane L. Abdoulaye (Niamey)
Demonstratives and their reference continue to generate a strong interest in linguistics research, being the subject of crosslinguistic or typological studies (cf. Anderson and Keenan 1985; Diessel 1999, 2003; Dixon 2003; Himmelmann 1996, 1997; etc.) and in-depth studies in particular languages (cf. Burenhult 2003; Enfield 2003; Lenz 2003; Morel/Danon-Boileau 1992; Weissenborn/Klein 1982; etc.). In Hausa studies, the locative adverbs (and their associated demonstrative pronouns and adjectives) have been the subject of repeated attempts at a characterization (cf. Gouffé 1971; Newman/Newman 1977: 18, 14; Newman 1990: 122, 276; etc.; cf. Jaggar/Buba 1994: 388–391 for the latest account and an overview of the previous accounts). Hausa is usually cited as having the following four locative demonstratives and their spatial reference: nân (or nàn) 'here', nan 'there near you', cân (or càn) 'there away from you and me', and can 'over there away from you and me' (cf. Jaggar 2001: 645; Newman 2000: 36–38). The referential meaning of these particles, at a first sight, seems indeed unproblematic, as seen in the following:
As suggested by the translations, nân refers to the position of the speaker, nan to the position of the hearer, cân to a position nearby but away from both speaker and hearer, while can refers to a remote location. Jaggar/Buba (1994: 391) interprete this referencing pattern as a person-oriented deictic system. In their analysis, nân would be associated with the 1st person, nan with the 2nd person, and cân and can with the 3rd person (following a model proposed in Lyons 1968: 278f.; cf. also Anderson/Keenan 1985: 282, 284). This account of Hausa dexis is now standard and is adopted in, for example, Jaggar (2001: 645), Newman (2000: 36–38), and Wolff (1993: 119). However, it is rather easy to find non marginal uses of the four adverbs that do not fit in the standard account, as seen next (cf. also Abdoulaye 2005: 85f.):
We see that on the one hand, examples (2) contrast a can and a nan that both refer to the position of the hearer. Indeed, the only difference between the two sentences is that in (2a) the speaker wants the hearer to stay away, while in (2b) the hearer is ordered to stand still. This shows that can does not code remoteness from both speaker and hearer. We also see on the other hand that in (3), nân can code the position of the speaker (if he/she is next to one end of the farm) or it can code a distal location (if the speaker is away from the farm's limits). In (4), nan and can code, respectively, a remote and a very remote location. Clearly, these facts, relating to the spatial use of the adverbs, cannot be explained by the standard account or in fact by any account that tries to map the reference of the four adverbs in a single, person or distance-based linear system.
The aim of this paper is to present a more exhaustive account of the deictic reference of Hausa adverbs nân, nan, cân, and can (and a fifth adverb ca̋an discussed in Section 3). It will in particular be shown that in order to understand the seemingly multiple values of the adverbs, as illustrated in (1–4), one needs to posit two deictic systems, both marked (nearly) with the same set of adverbial particles. The first system is based on a distance parameter while the other is based on a visibility parameter. Both systems, however, are complex in that the distance system embeds a person-based subsystem, while the visibility system embeds a distance-based subsystem.
The paper primarily has a descriptive concern and relies in various ways on recent typological and theoretical works in the domain of deixis, such as Anderson/Keenan (1985), Diessel (1999, 2003), Dixon (2003), Fillmore (1997), Himmelmann (1996, 1997), and the papers in Lenz (2003). Many typologies of demonstrative uses have been proposed (for a review see Himmelmann 1996: 218 and Imai 2003: 7). For example, Himmelmann (1996) distinguishes four major usage types for demonstratives. In the situational use, the reference of the demonstrative is located in the speech setting. This use can be illustrated in a sentence such as bring me that book!, where the book in question might be shown with an accompanying pointing gesture (cf. also Noonan 2001: 178). The situational use is thought by many authors to constitute the basic function of demonstratives, from which all other uses would be derived (cf. Diessel 1999: 110, Dixon 2003: 62–63, Fillmore 1997: 103). However, Himmelmann (1996) shows that demonstratives have three other main uses which are as universal as the situational use: the discourse deictic use, the tracking (or anaphoric) use, and the recognitional use. In the discourse deictic use, the demonstrative element refers to a proposition, an event, or a time point in the discourse, as in our forefather's occupation was abandoned, after that, it became like this, where that refers to the preceding proposition (adapted from Noonan 2001: 178). In the tracking use, the demonstrative refers to a participant in the story as in I drove the car to the parking and left it there where there refers to parking (cf. Fillmore 1997: 63). In the recognitional use, the reference of the demonstrative must be identified by the hearer based on shared knowledge, rather than situational cues, as in I am wondering what happened to that internal phone we used to have (cf. Himmelmann 1996: 231). This paper will essentially be concerned with the situational use of the Hausa locative demonstratives and the discussion of their other uses is deferred to another project.
It should, however, be noted that even the situational use has many subtypes, as discussed in Himmelmann (1996: 219–224). In their core situational function, demonstratives "are primarily used to focus the hearer's attention on objects or locations in the speech situation (often in combination with a pointing gesture)" (cf. Diessel 1999: 2). The speech situation, however, may be atypical in many ways. For example, a speaker-proximal adverb can have a relative reference and indexes the exact position of the speaker, the area he is in, the town, etc. (cf. Fillmore 1997: 64–65, Himmelmann 1996: 223). Distal adverbs may also have a "symbolic" use, such as when there in the sentence is John there (where you are)? refers to the place of the hearer over the telephone (cf. Fillmore 1997: 63). Demonstratives may also have a subjective/emotional use where the speaker can use a proximal adverb to refer to a remote place in order, say, to minimize its distance in the conversational setting or use a distal adverb to express emotional rejection of a referent (cf. Dixon 2003: 91). This paper will not be concerned with all these special uses and will mainly concentrate on the situational use in a typical, immediate speech setting, tracking the indications of distance or visibility of referents from the deictic center (speaker position).
The paper is organized as follows. Section 2 presents the values of nân, nan, cân, and can in the distance system and Section 3 presents the values of the adverbs nân, nan, cân, can, and ca̋an in the visibility system. Finally Section 4 gives some further evidence in support of a visibility system separate from the distance system.
2 The distance system
Jaggar/Buba (1994) give the most extensive account of Hausa deixis where they propose that the system is based on the person factor. In this regard, Jaggar (2001: 645) is even more explicit when he states:
This section takes another perspective and proposes that Hausa has a simple distance system which only secondarily involves a person subsystem in a structured, hierarchical reference scheme.
2.1 The basic proximal-distal contrast: nân vs. can
In data (2–3) we saw that there is a contrast between proximal nân 'here' and distal can 'there'. In this basic contrast, there is no reference to the hearer position and the combined references of nân and can cover the entire deictic domain around the speaker. The central thesis here is that can is a general distal marker that can refer to any location that is not speaker location. A simple evidence for the speaker-distal function of can is the fact that it is incompatible with verbs carrying the ventive maker -oo that codes some kind of motion towards the speaker. This is illustrated in the following:
In (5a), the basic form of the verb means 'go' and is compatible with can (it is also compatible with nân for reasons discussed in Section 3.1). However, in (5b), the verb is suffixed with the ventive extension -oo (cf. Newman 2000: 661–664) and only speaker-proximal nân 'here' is possible. In (5a), the first interpretation of can is 'there', i. e., a precise location. But in fact, can may also mean simply 'away from speaker'. For example, a speaker can use (5a) without a specific place in mind to signify to the hearer to simply go away (i. e., without a pointing gesture or with a "go away" gesture). However, even if a precise position is specifiable, the intended meaning of can can still be 'stay away from me' and in that case, the hearer of (5a) can stay still or go away. Another illustration of the general 'go away' meaning of can can be seen in the following (adapted from a radio broadcast play):
In the play, the reply in (6) was said by a wife to her husband who had just expressed a bit too much concern for the celibacy of Mariama. The wife then basically wishes that Mariama marry away from her husband. For this reason, can in this section is glos sed as "DthereG" for 'distance system general distal 'there'' and the nân glossed as "Dhere" for 'distance system proximal 'here''.
As a general distal marker, can can refer to any location as long as it can be conceived as being away from the speaker. For this reason, can can refer to very close or very remote and invisible places, as in the following (cf. also Abdoulaye 2005: 85):
Sentence (7a) might be said by someone reclaiming a bus seat. In this context, can would simply refer to the next seat where the hearer should move. In (7b) however, the sentence can be uttered, say in Africa, in which case can clearly refers to a very remote place. Next, we look at another contrast in the distance-based system, the person contrast.
2.2 The person subsystem: nan and cân
Although the general distal can may refer to the position of the hearer, Hausa has a special adverb nan which in the basic distance system refers to hearer location. This is already illustrated in the contrast in (2), where sentence (2b) with nan instructs the hearer to stand still, while (2a) tells him/her to stay away from speaker. In this function, nan contrasts with nân and cân, as seen in the following:
In (8), the most natural interpretation is where nân, nan, and cân refer, respectively, to the position of the speaker, the position of the hearer, and the position of a third party away from both speaker and hearer (but in their vicinity). In (8), the place specification (wuriinaa, wurinkà, etc. 'my place, your place, etc.') is not obligatory for all sentences to get the right interpretations (though a pointing gesture may be necessary for (8c)). To reflect their values, nan in this section is glossed as "Dthere2" for 'distance system hearer-centered 'there'' and cân glossed as "Dthere3" for 'distance system 3rd person-centered 'there''.
The most difficult problem for the hypothesis of a hearer-centered nan, independently of the value of nan in the visibility system (cf. Section 3.2), is the fact that there are very frequent uses of nan where the adverb does not refer to the position of the hearer. Indeed, without a place specification, (8b), i. e., lìttaafìi yanàa nan, can also have the quasi existential meaning 'the book is around'. Indeed, the adverb nan, besides the hearer-centered sense, can also refer to a certain known or presupposed location, the vague location of 'around, in the vicinity, etc.', or even simply to some place, as illustrated next (cf. also Abdoulaye 2006: 1153):
Among the four adverbs, nan alone may be used without a specific location to mean 'somewhere around'. In the first interpretation of (9a), the speaker says both that Kandin is at his known usual place (which can actually be specified) and alive. In the second interpretation, Kandin is said to be somewhere around or at some place and alive. In this existential-like interpretation, nan seems to have no reference to a specific location. Nonetheless, sentence (9a), even in its existential interpretation, still specifies that the referent is relatively "available" for the purposes of the participants in the conversation. Abraham (1959: 55), cited in Jaggar/Buba (1994: 388), has suggested a similar proposal and translated the example dookìi yanàa nan as 'a horse is available'. For this reason, sentence (9a) still differs from a plain existential sentence such as (9b) where the idea of relative availability to conversation participants is not an issue.
A good indication that the uses of nan illustrated in (9a), despite having something to do with location, are not deictic is the fact nan can refer to a position coded by nân (or another adverb) or otherwise presupposed in the same sentence, as seen in:
In (10a), nan has no independent deictic reference, and ultimately refers to the same speaker location referred to by nân 'here'. In (10b) also, nan, at some level, refers to speaker location, as suggested by the verb tàhoo 'come', which indicates that the object to be taken is close to the speaker. Taking together sentence (9a) and sentences (10), we see that nan – when it does not have the hearer-centered reference – can in fact refer to any location, including the location of the speaker. For this reason, this paper proposes to distinguish a (quasi) existential nan and a related general locative marker nan (glossed, respectively, as 'thereXST' and 'thereGL'), which would account for the double interpretation of (9a). The same two particles can also be seen in examples (10a) and (10b), respectively. Examples such as (10) contradict Jaggar/Buba's (1994: 399f.) claim that nan can be used to "pick out referents in non-spatial (anaphoric, symbolic) contexts" so as "to codify a more generalized speaker-distal reading", where the referent is closer to the speaker than would be indicated by DS can. Indeed, based on the crosslinguistic finding cited in Anderson and Keenan (1985: 287) that medial adverbs in languages tend to be used as "general anaphoric elements", Jaggar/Buba (1994: 400–408) propose that nan can have an anaphoric/cataphoric and symbolic use, as illustrated in the following (adapted from Jaggar/Buba 1994: 402, 405; cf. also Jaggar 2001: 646):
The typical anaphoric structure would be illustrated in (11a), where nan would refer to hantàa 'liver'. The second example would illustrate the symbolic use of nan, where we do not have a typical anaphoric structure (although in a similar example in Jaggar 2001: 646, nan is analyzed as cataphorically referring to the following specified location). There are, however, indications that these special uses of nan cannot simply be characterized as anaphoric or symbolic reference. Let us consider the following contrast:
In (12a–b), both nan and can are in a typical anaphoric construction and refer to Dan Isa. However, sentence (12a) with nan has two interpretations. The sentence can be realis, i. e., the hearer is instructed to go to Dan Isa and from there continue to Jibiya. The sentence can also have a hypothetical meaning and refer to the possibility for someone to be able to go to Jibiya from Dan Isa (in this case the second person pronoun would have an impersonal value; cf. Anderson/Keenan 1985: 260). By contrast, in (12b) with can, only the realis reading is possible. It may be the case then that adverb nan has shed its deictic meaning so much that it has in some contexts turned into a general locative adverb with the meaning 'a place'. This general meaning can clearly be linked to the nan illustrated in (9a) where the adverb has a quasi existential reading and means "somewhere around". The evolution of nan can be diagramed as follows:1
Of the three distal adverbs (i. e., general distal can, nan 'there near you', and cân 'there away from you and me'), it is not clear why only nan took up the existential function. It is also not clear why a 'there'-adverb should be selected for this "somewhere around (here)" function and not, say, the speaker proximal nân 'here'. What is clear,however, is that we find similar situations in other languages. For example in French, the (medial) distal là 'there' has taken up the quasi existential meaning (cf. il est là 'he is around, present', or even demain, je ne suis pas là 'tomorrow I won't be here/available'; cf. also English there are two people here with me).2
One important difference between general locative marker nan and hearer-centered nan is that the latter can always be replaced by the general distal adverb can, with the same reference but different connotation. This is illustrated in the following:
In (14), both nan and can have the same reference (assuming the hearer is holding the referent). In (14a) with the hearer proximal nan, the speaker just wants the hearer to keep the object and the speaker may come back to it. In (14b) on the other hand, the speaker is unlikely to come back to the object again. By contrast, when the general locative marker nan is replaced by can or any of the other adverbs, the meaning difference is more important. For example, nân may seem to be able to replace nan in examples (10) but the meaning will not be the same, as seen in the following for (10a):
Sentence (10a) can be used if the referent lives or works at the place and when the visitor came, the referent happened not to be around. In (15) however, it is understood that the referent does not live or work at the place, i. e., the speaker is looking for the referent at various places where she might go.
To summarize, if the existential and general locative functions of nan are put aside, the deictic reference in the distance-based system can be represented as in Table 1.
Table 1: The distance system (with an embedded person system)
In the distance system in Table 1, the fundamental proximal vs. distal contrast is speaker-oriented, i. e., nân is speaker-proximal and can speaker-distal. As a secondary contrast, the distal domain can further be specified along the person dimension where reference can be made to hearer position (second person) with the adverb nan or to a third party position (third person) with cân. The next section describes the visibility system.
3 The visibility system
Although the current standard account rejects it, the role of visibility in Hausa deixis has in fact been considered in previous studies. For example, Gouffé (1970/71: 299), cited in Jaggar/Buba (1994: 388), proposes that the falling tone adverbs nân and cân code visible referents, while the high tone adverbs nan and can code non-visible referents. Newman/Newman (1977: 18, 94) and Newman (1990: 122, 276), cited in Jaggar/Buba (1994: 389f.), proposes that while nân and cân code the visible referents, the high tone nan and can can code visible or invisible distal referents. Finally, Abdoulaye (1992: 239–240) states that Hausa has deictic markers "such as nân 'here', nan 'there (near addressee)', cân 'there (visible)', and can 'there (invisible)'". It should be noted that even Jaggar/Buba (1994: 391) also sensed in some way that their person-based system cannot be the whole story and write:
However, Jaggar/Buba (1994) did not systematically explore this cross-cutting system, nor do they recognize its organizing factor, i. e., the visibility factor, as suggested by previous writers.
This section argues that Hausa has a second deictic system based on the visibility factor. In this system, the entire visible field in front of the speaker is mapped by five adverbs, the already familiar nân, nan, cân, and can, and the less known ca̋an, i. e., a variant of can with an extra high tone and a long vowel. The section first discusses the comfortably visible area, mapped by nân and cân, and then the area visible only with some difficulty, mapped by nan, can, and ca̋an.
3.1 The comfortably visible area: nân and cân
It may seem a priori difficult to argue for a set of adverbs nân and cân coding visibility and different from the set nân and cân found in the distance system (DS). There are, however, some grounds for distinguishing a DS nân referring to a domain centered on the speaker from a visibility system (VS) nân referring to the visible area immediately in front of the speaker.
A simple evidence for two nân adverbs is the fact that nân can appear both with basic centrifugal verbs and –oo marked centripetal verbs. This is illustrated in the following:
As we saw in the discussion about data (5b), DS nân is naturally compatible with ventive verbs expressing some kind of motion towards the speaker. This is again illustrated in (16b), where the hearer is instructed to advance near the speaker. However, in (16a) nân appears with a basic motion verb that implies motion away from the speaker. The hearer of (16a) indeed will have to move away from the speaker to comply with the instructions. Since nân in (16a) does not refer to the location of the speaker, a pointing gesture is necessary to identify the place in question. That place must, however, be immediately in front of the speaker. This also applies to (3), where indeed a pointing gesture is necessary to indicate the nearest end of the farm. Let us also consider the following dialogue in (17) (adapted from Jaggar/Buba 1994: 398) and the example in (18) (adapted from Hiskett 1971: 76, with original translation):
Sentence (17b) shows that a speaker can use nân to reply to a question over a destination. Nonetheless, the place must normally be visible and identified with a pointing gesture (without a gesture, (17b) will have a euphemistic usage, where the speaker doesn't want to disclose exactly where he/she is going). In (18), the speaker explains that he/she lost a money bag nearby (cf. kusa 'in the vicinity').
In the visibility system, nân contrasts with cân, which also codes a comfortably visible reference but more distant from the speaker (i. e., not in his/her immediate vicinity). Cân is very likely the simplest of all Hausa demonstratives. According to Jaggar/Buba (1994: 409), cân is used for situational reference only (contrary to the other adverbs that can have a discourse or a tracking use). In its situational use, it also has about the same value in the distance system as in the visibility system. Indeed, VS cân codes a visible area which is contiguous to the first ground of the scene and hence still comfortably visible. This is already illustrated in (3), where the first end of the farm is referred to by nân and the second end referred to by cân. Nonetheless, VS cân is an integral part ot the visibility system, and, as seen later in Section 4, VS nân and VS cân pattern together in various respects and contrast with the other three adverbs in the visibility system.
3.2 The area visible with some difficulties: nan, can, and ca̋an
One particularity of the visibility system in Hausa is that all adverbs refer to visible locations. Indeed, the primary contrast concerns areas comfortably visible (nân, cân) and areas that are visible with some difficulty (nan, can, ca̋an). However, such a scale of visibility naturally translates easily as a distance scale. For this reason, early writers have indeed primarily detected the medial distal value of nan (and so missing its hearer-centered value). For example, Abraham (1959), cited in Gouffé (1971: 172), translates dookìn nan as 'the horse over there'. The three adverbs nan, can, and ca̋an are typically used when the speaker draws the hearer next to him and aligns his index finger before his eyes (usually with the arm fully extended) to indicate something far off. The adverb nan is used for the closest far off reference, can for the next step and finally ca̋an for the extreme, barely visible reference. The use of nan and can is already illustrated in (4), where Abdu, referred to with nan, is closest to the speaker and Musa, referred to with can, more remote. That can is indeed more remote than nan is shown by the fact that in a corrective localization, can must be used if the reference is further away than the initial localization. This is illustrated in the following:
In (19a), a tentative reference is set by Speaker A using nan. But the relevant referent happens to be even more remotely located. Speaker B, by using can, instructs Speaker A to jump over the first localization and search further away. Conversely, if the referent is assessed further than it is in reality, then the corrective localization will use nan to signify the referent is less remote. Jaggar/Buba (1994: 409) also noted the far off localization of can, as illustrated next:
According to Jaggar/Buba, the referent Abdu in (20) is at the limit of visibility since the description of his outfit stresses the difficulty of seeing him. Naturally, if Abdu's outfit style can be recognized, then clearly he would not literally be "at the limit of visibility" but simply at a remote distance and not easily identifiable. To reflect their values, nan in this section is glossed as "VthereLR" for 'visibility system less remote 'there'' and can glossed as "VthereR" for 'visibility system remote 'there''.
VS can differs from DS can in many respects. DS can is a general distal marker that can refer to locations in the vicinity of the speaker or locations very remote from him/her. VS can is far more restricted. Its reference is limited to locations that are remote from speaker but still visible. In fact there is a fifth adverb ca̋an, which previously has not been fully identified, and which refers to the extreme area of visibility, i. e., beyond the reference of nan or can but still visible. Formally, this adverb is a variant of can with an extra high tone and a long vowel (a peculiar syllable structure since normally Hausa syllables can only be CV, CVV, or CVC; cf. Newman 2000: 403). Ca̋an, for example, is used to refer to objects that, if they are moving away, are about to disappear from sight because of distance or because they reached the horizon line. In this way, it contrasts with VS can, as seen in the following:
In (21a) with ca̋an, the speaker is watching the referent becoming more and more indistinct from the background or about to cross behind the horizon. In contrast, in (21b) with VS can, t he speaker is watching the referent far off and about to disappear behind some obstacle. Although they do not formally identify it, Jaggar/Buba (1994: 409f.) do mention an "extreme/terminal" adverb can, which can be used to refer to the end of bounded dimensions such as the extremities of a caravane, the bottom of a (very deep) well, the top of a (very hight) tree, building, or structure, etc., as illustrated in the next data (adapted from Jaggar/Buba (1994: 411):3
In this example, it is understood that the climber reached the maximal height of the tree one can climb. The extreme/terminal use reported by Jaggar/Buba can be understood as an extension of the basic function of ca̋an, that of referring to the extreme limit of the visual field, typically the horizon. It may be noted that ca̋an can also be used for other perceptual modes such as hearing, as illustrated in the following:
In (23), the speaker describes the perception of a sound at the limit of the audible. For (23) to be felicitous the noise need not be far away, in fact it is typically close. This shows that ca̋an is used to map the extreme of a perception scale and not necessarily the physical distance of the noise. Ca̋an also seems to be the only adverb that can be reduplicated for intensification purposes, as seen in the next data:
Sentence (24) may be used to stress the difficulty of reaching, say, an animal that is hiding in the back of a room. Typically, the room need not be very long, but ca̋an can be used if the room is crowded and it is difficult to get to its back wall. To reflect these uses, ca̋an is glossed as "VthereXT" for 'visibility system extreme-terminal 'there''.
To summarize, besides the distance system, Hausa also has a deictic system based on visibility. This system uses five adverbs that can be represented with their reference as in Table 2.
Table 2: The visibility system (with an embedded distance system)
In Table 2, the falling tone nân and cân code the area comfortably visible to the speaker. The other three adverbs code the area visible only with some difficulty. Hausa is thus different from other languages reported to have a deictic system using visibility but which code visible vs. invisible referents (Dixon 2003: 90). It is also clear, however, that Table 2 also embeds a distance subsystem. The comfortably visible area is divided into a foreground marked by nân and a medial area marked by cân. The area visible with difficulty is divided into three: a less remote area marked by nan, a remote area marked by can, and an extreme area marked by ca̋an.
4 Further evidence for the role of visibility in Hausa deixis
The two previous sections described a distance and a visibility system for Hausa and one may doubt whether the two systems are really distinct. A possible alternative analysis can be found in Anderson/Keenan (1985). Anderson/Keenan (1985: 289–295) discuss languages that code more than one dimension in their deixis system, usually distance or person and another secondary dimension. However, in all the cases they discuss, the dimensions are clearly combined (for example 'near speaker + visible', 'near speaker + invisible', etc., for Kwakwa'la). Furthermore, in most cases, this combination is reflected in the complex morphology of the adverbs. There are no such combinations in the Hausa system. A second alternative analysis can be found in Jungblut (2003). According to Jungblut (2003: 28), languages can, so to speak, reorganize the reference system of their demonstratives depending on the spatial configuration of the conversation participants relative to each other. She shows for example that in Spanish, the deictic elements have different reference in face-to-face than in side-by-side communication settings. Clearly, one may seek to account for the Hausa seemingly double system along the lines of Jungblut's (2003) proposal, since the two systems share nearly the same set of markers. For example, a potential weak point in the double system analysis is the fact that some languages can extend the reference of their proximal marker to include the space in front of the speaker (cf. Fillmore's (1997: 121) example: Is this Harry Schwartz?, said while the referent stands meters away from the speaker).
All these alternatives notwithstanding, this section shows that besides the reference interpretations, there is further evidence for taking the visibility system as a basic system in Hausa deixis. In this regard, the section discusses two points, i. e., the common patterning of nân and cân and the role of the verb ga/ganii 'see' in Hausa deixis.
4.1 The priviledged formal relationship between nân and cân
According to most researchers (cf. Gouffé 1971: 170; Jaggar/Buba 1994: 414, note 1), the use of the four adverbs nân, nan, cân, and can is consistent throughout Hausa dialects. It is also recognized, however, that the western dialects may have variants that may or may not be formally related to the standard adverbs. For example, Katsinanci and most western dialects have variants for nân and cân but not for nan and can, as shown in the following (cf. Caron 1991: 70 and Mijinguini 1993: 65, 318, 320):
As one may see, only nân and cân have longer forms, from which they probably derive (Hausa has a rule turning long 'o' and 'e' into a short 'a' in closed syllables; cf. Jaggar 2001: 10, Newman 2000: 402). One may ask why nân and cân should in this way contrast with nan and can. In the linear distance system proposed in Jaggar/Buba (1994), the four adverbs have the following reference: nân 'here', nan 'there near you', cân 'there away from you and me', and can 'over there away from you and me'. In this system, there is no obvious reflection of the common pattern of nân and cân since they have no common function contrasting with the other two adverbs. Similarly, in the distance system proposed in Table 2, there is no obvious functional contrast between nân and cân on the one hand and nan and can on the other hand. However, if the visibility system is considered, one can cay that nân and cân, and their longer forms, map the comfortably visible area in front of the speaker. It should also be noted that Katsinanci has a formally unrelated variant for cân, which also has a short and a long form, as seen in the following (cf. Gouffé 1971: 174f.):
In Katsinanci, în has exactly the same uses as cân and its common formal pattern with nân and cân is hence not surprising. Besides the longer form, nân and cân are also used together in a fixed expression that can translate as 'here and there', as illustrated next:
In examples such as (27), the expression nân dà cân refers naturally to a few locations that typically were seen from a comfortable distance by the witness of the events. Nân dà cân is a fixed expression and, to our knowledge, there is no other such fixed expression combining two different adverbs (a temporal expression nan dà nan 'quickly' exists, which, however, involves the adverb nan only; cf. Jaggar/Buba 1994: 403).
To summarize, nân and cân have some formal relationship that can be well understood in the frame of the visibility system.
4.2 Verb ga/ganii 'see' in Hausa deixis
In western dialects, there is another deictic particle ga (cf. Gouffé 1971: 172), which functions primarily as demonstrative adjective or pronoun and locates referents in the visible proximity of the speaker. This subsection shows that there are good reasons to believe that this particle is related to the verb ga/ganii 'see'.
First, to better apprehend the functions of ga as a deictic particle, one may need to look at the function of nân, nan, cân, can, and ca̋an as demonstrative adjectives and pronouns. Indeed, all five adverbs can accompany a noun as demonstrative adjectives or combine with the indefinite pronoun wàa 'one' to form demonstrative pronouns or adjectives. These functions are illustrated for nân and nan in the following (cf. also Jaggar 2001: 324 and Newman 2000: 147):4
In (28), the adverbs accompany a noun, suffixed with the definiteness marker ‑`n, and function as demonstrative adjectives, while in (29) they form demonstrative pronouns with the indefinite pronoun wàa 'one', also suffixed with the definiteness marker. In (30), this demonstrative pronoun can in turn accompany a noun and so function as a demonstrative adjective, alternating with the simple adverbs, as seen in (28).
The particle ga, unlike the adverbs nân, nan, cân, can, and ca̋an, does not function as a locative adverb. However, like nân, nan, cân, can and ca̋an, ga can accompany a noun or combine with the indefinite pronoun wàa 'one' or the manner adverb hakà 'so', as illustrated in the following:
In (31a), ga acts like a demonstrative adjective and alternates with standard nân, nan, cân, can, and ca̋an, as seen in (28). In (31b), ga combines with the indefinite pronoun wàa 'one' to form demonstrative pronouns. Here, too, ga alternates with standard nân, nan, cân, can, and ca̋an, as seen in (29). Finally, in (31c), the demonstrative pronoun wânga 'this one' can function as a demonstrative adjective and accompany a noun, just like demonstrative pronouns formed with nân, nan, cân, can, and ca̋an. In all expressions of (31), the referent must be clearly visible in the situational use. For example, Gouffé (1971: 175) says that wânga 'this one' is equivalent to wannàn 'this one'. However, there is clear evidence that ga in fact refers to the comfortably visible area in front of the speaker. Indeed, the particle ga can combine with some of the standard adverbs, but not with others, as illustrated in the following:
In (32), ga combines only with nân and cân, a pattern that is best explained in the frame of the visibility system, i. e., the comfortably visible area in Table 2 is the reference domain of ga. In all probability, the deictic particle ga derives from the verb ga/ganii 'see' (the form of the verb depends on the syntactic context, cf. Newman 2000: 672). Indeed, it is generally considered (cf. Abraham 1959: 35, Newman 2000: 181) that the imperative form of the verb ga/ganii 'see' is the source of the presentational particle gàa '(t)here is/are' (cf. gàa Abdù '(t)here is Abdu', lit. 'see Abdu!'; gàa shi '(t)here he is', lit. 'see him!'). It happens that in the most westernly dialects of Hausa, the demonstrative pronoun wânga 'this one' alternates with an expression wân gàa shi 'this one', which combines the indefinite pronoun wàa and the presentational gàa shi '(t)here he/it is' in a demonstrative construction.
To summarise, there is indeed some evidence that Hausa has a genuine deictic system based on visibility that is separate from the distance/person system that mostly attracted the attention of researchers in Hausa studies.
This paper proposes that Hausa has in fact two separate deictic systems that nearly share the same adverbs: a distance system and a visibility system. The distance system has a fundamental speaker‑oriented proximal vs. distal contrast marked by nân 'here' and can 'there, away'. These two adverbs cover the entire deictic area around the speaker. There is, however, in the distance system a second contrast associated with a person subsystem and marked by nan 'there near you' and cân 'there away from you and me'. The second major deictic system in Hausa is the visibility system, which has five adverbs that taken together cover the entire visible area in front of the speaker. The fundamental contrast is between a comfortably visible area (mapped by nân and cân) and an area visible only with some difficulties (mapped by nan, can, and ca̋an). The paper shows that besides the reference interpretation of the adverbs, there is further evidence supporting the double system analysis. Hausa then shows that a factor other than distance can organize a deictic system.
Abdoulaye, Mahamane L. (1992): Aspects of Hausa morphosyntax in Role and Reference Grammar. Ph.D. dissertation, State University of New York, Buffalo.
Abdoulaye, Mahamane L. (2005): "A Review of Jaggar (2001): Hausa". Journal of African Languages and Linguistics 26: 81–120.
Abdoulaye, Mahamane L. (2006): "Existential and possessive predications in Hausa". Linguistics 44/6: 1121–1164.
Abraham, Roy Clive (1959): The language of the Hausa people. London.
Anderson, Stephen R./Keenan, Edward L. (1985): "Deixis". In: Shopen, Timothy (ed.): Language typology and syntactic description. Vol. 3: Grammatical categories and the lexicon. Cambridge: 257–308.
Beik, Janet (1987): Hausa theatre in Niger: A comtemporary oral art. New York/London.
Burenhult, Niclas (2003): "Attention, accessibility, and the addressee: The case of the Jahai demonstrative 'ton'". Pragmatics 13/3: 363–379.
Caron, Bernard (1991): Le haoussa de l'Ader. Berlin.
Diessel, Holger (1999): Demonstratives. Form, function, and grammaticalization. Amsterdam. (= Typological Studies in Language Series 42).
Diessel, Holger (2003): "The relationship between demonstratives and interrogatives". Studies in Language 27/3: 635–655.
Dixon, Robert .M. W. (2003): "Demonstratives: A cross-linguistic typology". Studies in Language 27/1: 61–112.
Enfield, Nicholas J. (2003): "Demonstratives in space and interaction: Data from Lao speakers and implications for semantic analysis". Language 79/1: 82–117.
Fillmore, Charles J. (1997): Lectures on deixis. Stanford. (= CSLI Lecture Notes Series 65).
Gouffé, Claude (1970/71): "Une corrélation typologique dans quatre langues de l'Afrique occidentale: les fonctions de *N-* )". Afrika und Übersee 54: 286–302.
Gouffé, Claude (1971): "Remarques sur le syntagme demonstratif en haoussa". In: Six, Veronica et al. (eds.): Afrikanische Sprachen und Kulturen – Ein Querschnitt. Hamburg: 170–179.
Greenberg, Joseph H. (1985): "Some iconic relationships among place, time, and discourse deixis". In: Haiman, John (ed.): Iconicity in syntax. Amsterdam: 271–287.
Himmelmann, Nikolaus (1996): "Demonstratives in narrative discourse. A taxonomy of universal uses". In: Fox, Barbara (ed.): Studies in Anaphora. Amsterdam: 205–254.
Himmelmann, Nikolaus (1997): Deiktikon, Artikel, Nominalphrase. Zur Emergenz syntaktischer Struktur. Tübingen.
Hiskett, Mervyn (1971): "The 'Song of the Shaihu's Miracles': A Hausa hagiography from Sokoto". African Language Studies 12: 71–107.
Imai, Shingo (2003): Spatial deixis. Ph.D. dissertation, State University of New York, Buffalo.
Jaggar, Philip J. (2001): Hausa. Amsterdam/Philadelphia. (= London Oriental and African Language Library Series 7).
Jaggar, Philip P./Buba, Malami (1994): "The space and time adverbials NAN/CAN in Hausa: Cracking the deictic code". Language Sciences 16: 387–421.
Jungbluth, Konstanze (2003): "Deictics in the conversational dyad: Findings in Spanish and some cross-linguistic outlines". In: Lenz, Friedrich (ed.): Deictic conceptualisation of space, time and person. Amsterdam/Philadelphia (= Pragmatics & Beyond 112): 13–40.
Lenz, Friedrich (ed.) (2003): Deictic conceptualisation of space, time and person. Amsterdam/Philadelphia (= Pragmatics & Beyond 112).
Lyons, John (1968): Introduction to theoretical linguistics. Cambridge.
Mijinguini, Abdou (1993): Karamin kamus na Hausa zuwa Faransanci. [Elementary Hausa-English dictionary]. Niamey.
Morel, Mary-Annick/Danon-Boileau, Laurent (eds.)(1992): La deixis: Colloque en Sorbonne, 8–9 juin 1990. Paris.
Newman, Paul/Newman, Roxana Ma (1977): Modern Hausa-English Dictionary. Ibadan/Oxford.
Newman, Paul (2000): The Hausa language: An encyclopedic reference grammar. New Haven.
Newman, Roxana Ma (1990): An English-Hausa Dictionary. New Haven.
Noonan, Michael (2001): "The 'double demonstratives' of Chantyal". Linguistics of the Tibeto-Burman Area 24/2: 173–188.
Weissenborn, Jürgen/Klein, Wolfgang (eds.)(1982): Here and there: Cross-linguistic studies on deixis and demonstration. Amsterdam.
Wolff, Ekkehard H. (1993): Referenzgrammatik des Hausa. Münster.
* Hausa (Chadic) is spoken mainly in Niger and Nigeria. Some of the data in this paper are constructed example sentences the grammaticality of which has been checked with speakers of Katsinanci dialect and Standard Hausa. Other examples are adapted from previous publications, as indicated sometimes in the text. Still other examples are adapted from naturally occurring utterances heard by the author. The transcription follows Hausa standard orthography with some changes. Long vowels are represented as double letters, low tone as grave accent, falling tone as circumflex accent, and the extra high tone as a double acute accent. High tone is unmarked. Small capitals (b, d, k, r) represent glottalized/laryngealized consonants and the trilled [r]. Written 'f' is pronounced [h] (or [hw] before [a]) in Katsinanci.
The abbreviations are: 1, 2, 3 '1st, 2nd, 3rd person'; CPL 'completive'; cop. 'copula'; df 'definite'; DS 'distance system'; f 'feminine'; FUT 'future'; imp 'impersonal'; IPR 'imperative'; IPV 'imperfective'; m 'masculine'; NEG 'negative'; p 'plural'; RI 'relative imperfective'; RP 'relative perfective'; s 'singular'; SUB 'subjunctive'; VEN 'ventive'; VS 'visibility system'. In addition, for the deictic adverbs, the following interlinear glosses are used: Dhere 'DS nân'; DthereG 'DS general distal can'; Dthere2 'DS hearer-centered nan'; thereXST 'existential nan'; thereGL 'general locative nan'; Dthere3 'DS away from speaker and hearer cân'; Vhere 'VS nân'; VthereLR 'VS less remote nan'; VthereR 'VS remote can'; VthereXT 'VS extreme ca̋an'.
I ow thanks to the participants at a talk audience at the Department of African Studies, University of Leiden, March 2004, for their comments and questions, to two anonymous referees whose comments helped improve this paper, and to my informants for their help in grammaticality judgments. I am naturally responsible for all remaining errors. back
1 In fact, there is a further non deictic use of the general locative marker nan in examples such as the following (cf. also Gouffé 1971: 174 and Jaggar/Buba 1994: 402):
In this example, the expression anàa nan, which can be repeated, has the temporal meaning 'meanwhile, one day, at some point, day in day out, etc.' This use is very likely derived from the existential or general locative use (in fact Gouffé translates the expression as 'one was there'). back
2 In the last few years, a number of studies have proposed, that in some languages at least one demonstrative marker may not have a distance-encoding semantics (cf. Enfield 2003: 91), and that for some deictic systems, one can do away with the notion of distance altogether (cf. Enfield 2003: 115). For these studies, the distance or location inferences drawn by native speakers stem in fact from the interaction between the choice of a particular demonstrative and the specific speech setting. Some authors have in particular singled out addressee-centered demonstratives and proposed that they code neither distance nor location (cf. Greenberg 1985: 278, Burenhult 2003: 378). I will leave the implications of these proposals for the Hausa addressee-centered nan to future more detailed studies. back
3 Conversely, Caron (1991: 70) reports that can can be realized on an extra high register. It is very likely that he is referring to the adverb ca̋an, although he does not explicitly give it the extreme-terminal interpretation. back
4 It should also be noted that in western dialects, fall tone adverbs nân and cân may undergo a rule of tonal simplification if they are preceded by a high tone to become nàn and càn. For all dialects, the simplification rule is obligatory in the demonstrative pronouns in (29) (cf. Newman 2000: 147). back