The roles of Dissociative and (Non-)Completive morphology in structuring Totela (Bantu) narratives

Full text

1
The Roles of Dissociative and (Non-)Completive Morphology
in Structuring Totela (Bantu) Narratives*
Thera Marie Crane
University of Helsinki
Abstract
In Totela, infinitive-based ‘narrative’ morphology alternates with forms
inflected for tense and aspect. While narrative morphology can depict
sequential events, inflected forms are used with both non-sequential and
sequential event predicates. When inflected forms appear, especially in
contexts where narrative morphology might also be appropriate, they play
important roles in signaling narrative structure. The three most common
categories of inflected verbs in narratives indicate ‘completion’, ‘non-
completion’, and ‘dissociation’. Dissociative marking appears at the
beginning and ending of a narrative and frames it by shifting the cognitive
domain to a world, separate from the world of telling, where listener belief
is suspended to include narrative events. Inside that world, Completives
and Non-completives provide structure to the narrative and direct listener
responses.
*Many thanks to Cecilia Namasiku Namuyumba, Gertrude Sibeso, Violet Bumba,
Christopher Mwendo, Albert Mwenda, Phineas Simwaga Sishau, Gift Mwakamwi
Sishau, and Kelvin Sishau, and to other residents of Kwemba village, particularly to
headman Sishau White Maketu, for their extremely generous contributions to all aspects
of this research. Thanks also to the communities of Likemwa, Samisisi, and Malabwe in
Zambia, and of Makusi and Kachansi in Namibia, and to consultants and research
partners Michael Sililo, Veronika Kalimukwa, Clement Tubusenge, Tekulo Kachelo,
Samitiba Agatha Nasamu, Kelly Mutale, and Namasaka Imuwana. Work in Zambia and
Namibia was made possible through research affiliations with the University of Zambia
and the University of Namibia. Elicited material is based upon work supported in part by
a National Science Foundation Graduate Research Fellowship. Any opinions, findings,
conclusions or recommendations expressed in this publication are those of the author and
do not necessarily reflect the views of the National Science Foundation. Field research
was also made possible by grants from the UC Berkeley Graduate Division and the
Andrew and Mary Thompson Rocca Summer Pre-Dissertation Research Award in
African Studies (UC Berkeley), and by the Harvey Fellows program. Thanks also to the
editors of this volume and the two anonymous reviewers for their very careful readings of
an earlier version and their insightful and helpful comments. All errors, of course, are my
own. Portions of this paper are revisions of earlier work including Crane (2011a, 2011b,
2011c).
This pdf is the author accepted manuscript of the following copyrighted publication:
Crane, Thera Marie. The roles of dissociative and (non-)completive morphology in
structuring Totela (Bantu) narratives. In Doris L. Payne & Shahar Shirtz (eds.) Beyond
aspect: The expression of discourse functions in African languages, 145-176.
Amsterdam/Philadelphia: John Benjamins Publishing Company.
http://dx.doi.org/10.1075/tsl.109.06cra . Please contact the publisher for permissions to
reuse or reprint the material in any form.

Dissociative and (Non)-Completive Morphology in Totela Narrative
2
1 Introduction
As is the case in many Bantu languages (see e.g. Nurse 2008), the great
majority of verbs in Totela narratives are marked with narrative
morphology. Narrative morphology predictably appears on verbs
describing sequential events, and—as is also common—is formed in
Totela with a bare infinitive (cf. Carlson 1992: 82-83, who notes that
decreasing finiteness correlates with increasing “continuity”). Totela
narrative morphology is illustrated in example (1).1
(1) kú-kwààt-à kù-nèns-à kù-nèns-à
NARR-grab-FV NARR-beat-FV NARR-beat-FV
‘[then she] caught [him and] beat [him and] beat [him]’
(Kañandu)
Of greater interest in this article, however, is the use of verb forms that
are inflected for tense and aspect. These forms appear not only on verbs
describing non-sequential events, but are also used with some frequency to
mark sequential events, where narrative morphology might normally be
expected. This article deals with the uses of this ‘marked’ inflectional
morphology within Totela narratives, and the storytelling goals that
speakers fulfill when employing it.
The alternation of narrative and inflected forms can be thought of as a
kind of ‘tense switching’, for which Fleischman (1990) notes a variety of
functions, including marking shifts in “subject and/or of discourse topic”
(200), dividing a narrative “into spans centered around a setting or macro-
event” (201), helping with narrative “pace” (210-211), and shifting point
of view (216ff). Nurse further notes that when interrupted by non-
narrative forms, a sequence of narrative markers can be resumed “to stress
continuity” (2008: 120).
1 The practical orthography (PO) used in Totela examples is similar to that of siLozi (and
many other languages of Zambia). Significant departures from standard phonological
notation are as follows:
IPA
PO
IPA
PO
b
bb
ʃ
sh
β
b
tʃ
ch
ɲ
ny
ʤ
j
ŋ
ñ
j
y
hwu
hu (labialized, often nasalized glottal
fricative, occurring when glottal fricative
is followed by /u/)
Vowel length is indicated by vowel doubling when it is potentially lexically contrastive.
Some degree of vowel lengthening is also observed before prenasalized stops, but is not
contrastive and is not marked in the orthography. All tones (H and L) are exhaustively
marked on examples for which the tones have been verified in elicitation in order to
distinguish between all L utterances for which tone is known and utterances for which
tones have not been verified.

Dissociative and (Non)-Completive Morphology in Totela Narrative
3
In this article, I focus in particular on the three most common
categories of verbal inflection that appear within narratives: markers of
‘completion’, ‘non-completion’, and ‘dissociation’. The narrative uses are
seen to follow from the so-called ‘basic’ (i.e. non-narrative) meanings, and
a close examination of the categories’ uses in narratives in turn allows for
a fuller understanding of their overall nature. I argue that gaining such a
perspective requires us to move from purely temporal definitions of these
‘tense’ and ‘aspect’ forms to a broader, discourse-based, and—
ultimately—better analysis of their semantic and pragmatic functions.
The remainder of the article is organized as follows: Section 1.1
introduces the Totela language. Section 1.2 provides definitions of
terminology that will be used throughout the article and gives brief notes
about the orthography and other technical issues related to discussing the
narrative uses of TAM forms. These introductory notes are followed in §2
by general descriptions of the Completive, Non-completive, and
Dissociative forms analyzed in this article, along with their ‘basic’
functions in Totela. I argue in §2 that typical labels for tense and aspect
categories such as ‘perfective’ or ‘present’ serve to obscure rather than
expose the nature of Totela verbal forms, and that a different
conceptualization is necessary, both for understanding these forms in non-
narrative contexts, and for understanding how they are used in narratives.
Narrative uses of the forms and their alternations with narrative
morphology are explored in §3 (Completive and Non-completive forms)
and §4 (Dissociative forms). Section 5 concludes the article with a brief
discussion of the need for a holistic view of tense and aspect marking.
1.1 Totela
Totela is a highly endangered Bantu language spoken—in two
strikingly different varieties—in Zambia’s Western Province and in the
Caprivi Strip in Namibia (Crane 2011). It is classified as K41 in the
Guthrie classification system (see Maho 2009) and is part of the Bantu
Botatwe group, which comprises certain languages in the M and K zones,
including Tonga (M64), Ila (M63) and Subiya (K42) (see de Luna 2008
for further details). The true number of Totela speakers is unknown, but
Crane estimates around 3,000 Totela speakers in rural areas of Zambia’s
Western Province (2011: 56). Virtually no children learn Totela as a first
language in Zambia, so the Totela-speaking population in Zambia is
rapidly dwindling, and the variety may be considered moribund.
Data for this study are drawn from my corpus of narratives and
elicitation data collected during three research trips in 2006, 2007, and
2009. The article deals only with forms as they are used in the Zambian
variety of Totela, which exhibits significant phonological, morphological,
and tone differences from its Namibian counterpart (see Crane 2011 for

Dissociative and (Non)-Completive Morphology in Totela Narrative
4
details).
1.2 Terminology and conventions
The infinitive morphology used to mark consecutive events in narratives is
referred to as ‘narrative morphology’ (glossed as NARR).2 When an
infinitive form is used as, for example, a gerund or participle within a
verbal construction, it is glossed as an infinitive (INF).
Verbal prefixes of tense and aspect are written with dashes on both
sides of the morpheme if they follow the subject marker and precede the
root (e.g. -la-, as in ndi-la-neng-a ‘I dance’), or with standard prefix
notation if they precede the subject marker (e.g. na-, as in
na-ndi-la-neng-a ‘I will dance’). Although both are in fact verbal prefixes,
this notational convention allows for easily distinguishing between pre-
and post-subject-marker morphemes with the same segmental
morphology, such as prehodiernal Dissociative markers ka- (Imperfective)
and -ka- (co-occurring with Completive).
‘Perspective Time’ is defined following Condoravdi (2002) (along
with Cover 2010: 50) as the time from which a proposition’s truth value is
assessed. Although this time is typically equivalent to utterance time, it
may also be shifted. Condoravdi (2002: 62) gives the following example
from English:
(2) At that point he might (still) have won the game, but he didn’t in
the end (Condoravdi 2002: 62, ex 7).
From the first clause’s Perspective Time of ‘at that point’—which is
not equivalent to utterance time, but prior to it—it was true that the subject
had the possibility of winning the game. If utterance time were always
equivalent to Perspective Time, then the first clause would be necessarily
false.
Perspective Time is also sometimes referred to as ‘evaluation time’ in
the literature (see e.g. Hatav 2012, who defines ‘evaluation time’ as “the
orientation time of a clause” and points out that Reichenbach’s (1947)
‘reference time’ seems to encompass both evaluation time and ‘location
time’ (Hatav 2012: 611).
Setting Perspective Time to “time as it progresses through a
narrative” allows for the narrator’s use of non-past forms without
necessarily breaking out of the time of the narrative.
In comparing the use of narrative morphology and inflected verbs in
narrative, only non-quoted speech will be considered in this article.
2 It can be noted that Totela does not employ prototypical ‘serial’ verb constructions,
which might present difficulties in establishing narrative ‘events’. Narrative morphology
lacks some important characteristics of serial verb constructions. For example it can be
used across clauses, with changing subjects (whether overtly indicated or not), and it is
often linked by coordinating conjunction na-.

Dissociative and (Non)-Completive Morphology in Totela Narrative
5
Storytellers regularly use inflected verbs when speaking through the
voices of a character. The use of inflected verbs in quoted speech is
expected, as character speech mirrors everyday speech.
2 Dissociation and Completion in Totela
This section introduces the three tense/aspect categories that will be
discussed in §3 and §4 below. A shallow analysis of these categories
might lead to labeling them ‘perfective’, ‘present’ (or ‘non-past’) and
‘distal past’, but closer examination reveals that these labels do not
adequately describe their functions. Understanding them instead as
markers of ‘nuclear completion’ and ‘non-completion’ not only better
explains their temporal and semantic variations, but also illuminates their
narrative-structuring roles.
In §2.1 I discuss the notion of ‘nuclear completion’ (or non-
completion). This discussion requires an exposition of the basic situation
type (Aktionsart) categories in Totela (§2.1.1), and how they differ from
the Vendlerian categorizations typically applied to predicates in English
and many other languages. Sections 2.1.2-2.1.3 show how the main
situation types in Totela interact with verbal marking to produce
Completive and Non-completive temporal semantics.
Section 2.2 discusses Botne and Kershner’s (2008) notion of temporal
dissociation – which they argue constitutes true ‘tense’—and how it
manifests in Totela tense and aspect marking.
The discussion in §2.1-2.2 may be found in greater detail in Crane
(2013).
2.1 (Non-)Completion
2.1.1 Situation types
To understand the temporal semantics of (Non-)Completive forms in
Totela, it is necessary to recognize a basic distinction in Totela situation
types. Following Botne & Kershner (2000, who in turn draw ideas from
Beuchat 1966; Freed 1980; Botne 1983 and others), this article
distinguishes two main categories of situation type in Totela, here referred
to as ‘change-of-state’ and ‘durative’. These two situation types differ
primarily in terms of the situation ‘nucleus’, described by Botne &
Kershner as “constituting the characteristic and prominent feature of the
event” (2000: 165).
Specifically, a situation may have three phases: (optional) onset O,
(required) nucleus N, and (optional) coda C. The onset may include
preparatory phases, and the coda, result phases, either lexically-entailed or
otherwise.
Durative situations correspond roughly, though not precisely, to

Dissociative and (Non)-Completive Morphology in Totela Narrative
6
Vendler’s ‘states’, ‘activities’, and ‘accomplishments’. They do not
express a change from one state to another, and they have a non-punctual
nucleus. The nucleus coincides with the action or state described, and ends
upon termination of the situation. Durative situations can include non-
change-of-state statives (e.g. -saka ‘to want’; there are few such verbs in
Totela),3 activities (e.g. -samba ‘to bathe’), and other situation types that
do not encode a change of state. A simple event structure for durative
-samba ‘to bathe’ is schematized in (3).
(3)
[
bathing
]
(O)
N
(C)
Change-of-state situations, in contrast, have a nucleus that represents
the point of “change of condition or location of the experiencer or
patient”, potentially a “change or transition from one state to another”
(Botne & Kershner 2000: 165). These situations are sometimes referred to
as ‘inchoative’ and are somewhat analogous to Vendler’s ‘achievements’,
but entail a result state and are pervasively used to depict ongoing states.
An example of a change-of-state verb in Totela would be -komokwa
‘get/be surprised’, with the nucleus representing the change to a state of
surprise. Other examples include -bomba ‘soak’ and -ikuta ‘become full’.
The onset phase may or may not be a temporally-durative process. The
result states of these verbs (here, being surprised, being soaked, and being
full) cannot be expressed with a present-tense form. Rather, they require a
form locating Perspective Time in the post-nuclear state, as shown in (4–
5), which contrast Completive marking (a) with Non-completive forms
(b).
(4) a. ndá-kòmòk-w-à
1SG.CMPL-surprise-PASS-FV
‘I’m surprised’
b. ndì-lá-kòmòk-w-à
1SG-NONCMPL-surprise-PASS-FV
‘I’ll be surprised’
(5) a. òmwànjà wà-bòmb-à
CL3.cassava CL3.CMPL-soak-FV
‘the cassava is soaked’
b. òmwànjà ù-là-bòmb-à
CL3.cassava CL3-NONCMPL-soak-FV
‘the cassava is soaking / the cassava will soak’
In each pair, the contrast between (a) and (b) shows that a present stative
reading is obtained not with the Non-completive forms in (b), but only
3 Statives of perception, e.g. ‘hear/understand’ or ‘see’ can be used either as durative or
as change-of-state predicates in Totela.

Dissociative and (Non)-Completive Morphology in Totela Narrative
7
with the Completive forms in (a). The interactions of the Completive and
Non-completive forms with various situation types are discussed more
fully in subsequent sections. Note that in contrast, durative situations have
a past-tense reading with Completive marking, and a present or future
reading with Non-completive marking, e.g. ndà-sàmb-à (Completive, ‘I
(have) bathed’) and ndì-là-sàmb-à (Non-completive, ‘I (will) bathe’).
Example (6) gives a rough schematization of the change-of-state
situation -bomba ‘soak’.
(6)
in process of soaking
[
in state of being soaked
O
N
C
In this example, the nucleus represents the point at which the speaker
construes the subject in question to have passed from the process of
soaking to a state where can be considered to be ‘soaked’. The lack of a
closing bracket indicates that the end of the coda situation (being soaked)
is not necessarily a part of the situation's event structure.
Further distinctions can be made within the umbrella categories of
durative and change-of-state categories, and are discussed in Crane
(2011). For the purposes of the following discussion, the crucial
distinction is that between a change-of-state situation, which has a
punctual nucleus and a result-state coda, and a durative situation, which
has a non-punctual nucleus. It is important to note that—while details and
analyses differ somewhat—a distinction of this sort is prevalent across
many Bantu languages. For examples, see Fortune (1949), Botne (1983),
Botne & Kershner (2008), and Nurse (2008), among many others.
With these main situation-type distinctions in mind, I now turn to a
discussion of how they interact with Completive and Non-completive
forms, respectively.
2.1.2 Completive -a-
Morphologically, the Totela Completive is marked by an -a- marker
that appears immediately after the subject agreement marker, fusing with
it morphologically (e.g. 1SG ndi- becomes nda-; 1PL tu- becomes twa-).
With durative verbs, -a- gives a past or perfect reading.
(7) twà-nèng-à
1PL.CMPL-dance-FV
‘we danced’ / (contextually available:) ‘we have danced’
(8) ndà-w-à
1SG.CMPL-fall-FV
a. ‘I fell [and got back up again]’ / b. ‘I have fallen [and am still
lying on the ground]’

Dissociative and (Non)-Completive Morphology in Totela Narrative
8
With change-of-state verbs, the -a- marker usually has a result-state
implicature at Perspective Time, as in (9).
(9) ndá-kòmòk-w-à
1SG.CMPL-surprise-PASS-FV
‘I’m surprised’
However, the result state does not necessarily hold at Perspective
Time and can be defeased by particular lexical items. For example, (10)
begins a narrative about cows trampling some crops.
(10) ndá-kòmòk-w-á sùnù!
1SG.CMPL-surprise-PASS-FV today
‘I had a surprise today!’
Table 1 summarizes the possible readings of Completive -a- with durative
and change-of-state verbs.
Table 1. Available readings of Completive -a- by situation type
Durative
Change-of-State
perfective
(default reading)
present stative
(default reading)
perfect
(contextually-available
reading)
perfective
(contextually-
available reading)
With this variety in its possible interpretations, the -a- form cannot be
neatly described as either a perfective or an imperfective marker ‘viewing’
the situation as a whole or from the inside. Likewise, the label ‘perfect’ as
it is typically used does not fit this marker. As seen in the examples above,
the Perspective Time is not necessarily in the result state (note the possible
interpretations of (8) as well as example (10)). Furthermore, as in (7), the
-a- form is used to express general (hodiernal) pasts. This use would be
unexpected with true perfect forms, which additionally—unlike -a- forms
in Totela—are rarely found in narrative sequences (Labov & Waltezky
1967: 90; Dahl 1985: 138; Fleischman 1990: 30). Further arguments on
why -a- cannot be analyzed as a perfect are given in Crane (2011a,
2011c).
Rather than as a ‘perfective’ or as a ‘perfect’, -a- can best be
described as a ‘Completive’ form that locates Perspective Time
subsequent to the situation nucleus. Perspective Time can either fall within
the result state—if such a state exists—giving perfect and resultative

Dissociative and (Non)-Completive Morphology in Totela Narrative
9
readings such as (8b) and (9), or after the nucleus and any result state,
giving past/perfective readings as in (7), (8a), and (10).
Examples (11) and (12) give simplified schematizations for a durative
and a change-of-state verb, respectively. PT represents the location of
Perspective Time. The (a) examples in each pair illustrate the default
readings, while the (b) examples represent other contextually-available
readings.
(11) a. ndàsàmbà ‘I bathed’
(O)
N
[
bathing
]
PT
b. ndàsàmbà ‘I have bathed’
(O)
N
C
[
bathing
]
state of being clean
PT
(12) a. wàbòmbà ‘It is soaked’
O
N
C
in process of soaking
[
in state of being soaked
PT
b. wàbòmbà ‘It got soaked’
O
N
C
in process of soaking
[
in state of being soaked
]
PT
2.1.3 Non-Completives
In contrast with Completive -a-, Totela’s Non-completive forms do
not necessarily have any overt TAM morphology associated with them.
They can occur either as a verb unmarked for tense and aspect
(SM-root-a) or marked with the morpheme -la- following the subject
marker (SM-la-root-a).
Unmarked verbs and -la–marked verbs are each restricted to certain
morphosyntactic environments, which are described in Crane (2011). The
morpheme -la- is clearly related to a historical disjunctive marker (cf.
Güldemann 2003, Nurse 2008: 205ff, and Crane 2011) and still retains
some traces of its disjunctive functions. For the purposes of this article, the
differences between unmarked and -la-–marked verbs are not significant,
and they can be treated as interchangeable. It can be noted that the Non-
completive is not necessarily a form of marking per se, but rather—

Dissociative and (Non)-Completive Morphology in Totela Narrative
10
especially in the case of the unmarked form—the absence of Completive
morphology.
With durative verbs, these forms allow ongoing (stative or active)
readings (13–14), near-future readings (15), and habitual and generic
readings ((16) and (17), respectively). These readings depend upon the
context in which the verbs are used.
(13) obwawu bu-la-shup-a
CL14.relish CL14-NONCMPL-cause.trouble-FV
‘relish is a problem’ (i.e. hard to come by)
(14) kùnò kù-là-lìl-w-à
CL17(LOC).DEM CL17(LOC)-NONCMPL-cry-PASS-FV
‘there is mourning [going on] here’
(15) ìjìlò ndì-là-y-á kù-mpìlì
tomorrow 1SG-NONCMPL-go-FV CL17(LOC)-CL9.fields
‘tomorrow I’ll go to the fields’4
(16) munsikunsiku ndi-la-yend-a
every.day 1SG-NONCMPL-walk-FV
‘I walk every day’
(17) otuyuni tu-la-tus-a abantu
CL13.bird CL13-NONCMPL-help-FV CL2.person
‘birds help people’
With change-of-state predicates, on the other hand, present and
stative readings are unavailable with -la- (although habitual readings
are possible). Instead, a future interpretation is forced, as may be seen
in the contrast between the main clause predicates in (18–19). In (18),
the state ‘to be happy’ with change-of-state verb -taba ‘become happy’
must be expressed with a completive form. When -la- is used with a
change-of-state form (19), the reading of that is of a future state that
does not hold at Perspective Time.
(18) àbá bà-yîmb-à bà-tàb-à
CL2.DEM CL2-sing-FV.RC CL2.CMPL-become.happy-FV
‘those who are singing are happy [CMPL]’
(19) àbá bà-ly-â bà-lékùt-à
CL2.DEM CL2-eat-FV.RC CL2-NONCMPL.become.full-FV
‘those who are eating will be full [NONCMPL]’
4 The Non-completive is generally restricted to hodiernal futures, but when conveying
certainty or immediacy, can also be used in post-hodiernal contexts.

Dissociative and (Non)-Completive Morphology in Totela Narrative
11
Table 2. Available readings of Non-completive by situation type
Durative
Change-of-State
present (stative/active)
near-future
habitual/generic
future
habitual/generic
In contrast to the Completive, then, Non-completive forms seen to indicate
that Perspective Time is located prior to the completion of the situation
nucleus. Since change-of-state situations have a punctual nucleus, they
must receive a future reading with Non-completive forms. With durative
situations, the Non-completive can locate situation time either prior to
(near-future) or within (present activity/state) the situation nucleus.
Sample schematizations are given in (20) and (21). (20a) represents a
present reading with a durative predicate, with Perspective Time located
within the situation nucleus; while (20b) shows the same predicate with a
future reading, where Perspective Time is located within a contextually-
construed onset.
(20) a. ndìlàsàmbà ‘I bathe / am bathing’
(O)
N
(C)
[
bathing
]
PT
b. ndìlàsàmbà ‘I will bathe’
O
N
(C)
planning or preparing to bathe
[
bathing
]
PT
Example (21) shows two possible readings with a change-of-state
verb, with Perspective Time located respectively within the onset phase
(a) and prior to it (b).
(21) a. úlàbòmbà ‘It is getting soaked’
O
N
C
in process of soaking
[
in state of being soaked
PT
b. úlàbòmbà ‘It will (be) soak(ed)’
O
N
C
in process of soaking
[
in state of being soaked
PT
2.2 Dissociation

Dissociative and (Non)-Completive Morphology in Totela Narrative
12
In addition to Non-completive and Completive expressions, Totela
also has forms that mark situations as having obtained before or after the
day of Perspective Time. This section introduces these forms and argues,
following proposals by Botne & Kershner (2008), that they should be
analyzed as shifting discourse to domains that are cognitively dissociated
from here-and-now reality.
Botne & Kershner (2008; see also Botne 2012, Sweetser &
Fauconnier 1996) describe two domains, or ‘cognitive worlds’, which they
label the ‘P-domain’ and the ‘D-domain’, respectively. The P-domain (or
“primary” domain, as in Botne 2010) represents the world of the time,
place, and reality of the speech act (or Perspective Time); in contrast, the
D-domain (or “Dissociative” domain) marks “relations of non-inclusion”
in the here-and-now world of Perspective Time. (Botne & Kershner 2008:
153). The two domains are schematized in Table 3.
Table 3. Cognitive domains in Botne & Kershner (slightly adapted
from Botne & Kershner 2008: 159)
P-Domain: Association
= inclusion in world of
Perspective Time
D-Domain: Dissociation
= exclusion from world of
Perspective Time
Reality
real
not real
Time
now
not now
Space
here
not here
Languages may encode any or all of the oppositions in Table 3.
Interestingly, identical morphemes often encode more than one opposition.
For example, the English Past can either encode time (I ate the cookie) or
reality status (If I ate the cookie...).
When used without additional domain marking, the Completive and
Non-completive forms discussed up to this point situate discourse in the
temporally associative P-domain. They can be used to describe past or
future situations, but without additional tense/aspect morphology they
usually reference situations with a nucleus that is at least partially
contained within the day of Perspective Time. To refer to situations prior
to the day of utterance,5 Totela speakers use either the marker ka-,
5 It should be noted that the ‘day of utterance’ seems to be construed by speakers as
starting at the time of going to sleep—not, as is claimed to be the case for many
languages, at sun-up (see e.g. Nurse 2008: 90). This timing of ‘hodiernality’ is supported
by the claims of Dahl (2008, 2009), who cites studies at Brown University and the Max
Planck Institute in Heidelberg suggesting that memories are stored both permanently in
the neocortex and temporarily in the hippocampus. The hippocampus is “wiped” during
sleep each night. In Totela, then, it seems that experiences stored in both short-term and
long-term memory are expressed with Associative forms, while experiences stored only

Dissociative and (Non)-Completive Morphology in Totela Narrative
13
prefixed to the subject marker, and with an associated tone pattern (22), or
the marker -ka-, which follows the subject marker and the Completive -a-
morpheme (23). These prehodiernal Dissociative forms will be
collectively referred to as ka forms.
Ka-SM is used to describe situations from an imperfective viewpoint;
the SM-a-ka form, which indicates nuclear completion in a prehodiernal
domain, has a prehodiernal perfective (Completive) reading.
(22) kà-ndì-yênd-à
PREHOD.IPFV-1SG-walk-FV
‘I was walking / I used to walk’
(23) ndà-kà-yènd-à
1SG.CMPL-PREHOD-walk-FV
‘I walked (yesterday or before)’
Similarly, a prefix na- is used when referencing situations a day or
more after the time of utterance. In all contexts where the -la- marker
would appear in hodiernal Non-completives (e.g. in most affirmative
main-clause indicatives), it also appears in forms with na-.
(24) ná-ndì-là-yènd-à
POSTHOD-1SG-NONCMPL-walk-FV
‘I’ll walk (tomorrow or later)’
Several lines of evidence point to the ka and na- forms as
referencing a separate cognitive domain, rather than merely indicating a
certain time frame.
First, there is some flexibility in the use of ka-/-ka- and na- for non-
hodiernal time frames. Although neither can in general be used for
situations that do coincide with the hodiernal (today) time frame, neither is
universally required when reference is made to non-today situations: there
is at least some optionality in their use. For example, the construction
in long-term memory are expressed using Dissociative marking. This distinction in access
to memories and the corresponding marking could be further evidence in favor of the
existence of cognitive domains that align with ‘today’ and ‘not-today’ situations. Indeed,
the majority of languages that morphologically mark distinctions in temporal distance
encode the difference between ‘today’ and ‘not today’ (although other systems exist).
This distinction may be seen as primary to other temporal distance distinctions (see e.g.
Dahl 2008, 2009; Botne 2012: 537). Note, however, that for many languages with finer
grained remoteness distinctions (e.g. today vs. yesterday vs. distant past), it can be argued
that only the distant past represents a dissociative domain (see e.g. Botne 2012 for
examples).

Dissociative and (Non)-Completive Morphology in Totela Narrative
14
chìlìmò chàmánà (‘last year’; literally ‘year that has finished’), without
prehodiernal -ka-, alternates with -ka-marked chìlìmò chàkàmánà, also
meaning ‘last year’, with the former version seemingly indicating a time
period that is somehow more relevant to the present.
A discourse example of a prehodiernal situation not marked with -ka-
is given in (25).
(25) Chwale nje-yeyi nchechi mwa-waan-a
INTERJ COP.CL9-CL9.DEM CL9.church 2PL.CMPL-find-FV.RC
‘and so that’s the church you(’ve) found [CMPL] [here today],
Inyweb a-Thera. Mwa-tu-waan-a
2PL.PRON 2PL-Thera 2PL.CMPL-1PL.OM-find-FV
Thera. You(’ve) found [CMPL] us
mu-mapulogalama a-mangi.
CL18(LOC)-CL6.program CL6-many
in the midst of a lot of programs.’
Although Completive forms are generally used for hodiernal situations,
they are used in (25) to describe my arrival several weeks prior to the
utterance. However, the relevant information in the utterance is the
busyness of church life in the village, which had not changed since my
arrival and which continued to affect the availability of certain research
consultants. The arrival, while not in the precise temporal domain of
Perspective Time (i.e. the day on which the utterance was made), was still
within its cognitive domain: the busyness of church life had a direct effect
on the here-and-now world of the research situation. A similar example is
given in (26).
(26) twà-yènd-á àmázùbà ò-bílè, ndé-tù-sìk-à
1PL.CMPL-walk-FV CL6.day CL6-two DM-1PL-arrive-FV
‘we(’ve) walked [CMPL] two days, now we’re just arriving
[NONCMPL]’
The use of a hodiernal form in (26) situates the walking event within the
cognitive domain of the arrival.
Even more common is the description of posthodiernal situations
without the use of the na- marker. For example, both (27a) and (27b) are
perfectly acceptable.
(27) a. ìjìlò ndì-là-y-á kù-mpìlì
tomorrow 1SG-NONCMPL-go-FV CL17(LOC)-CL9.fields

Dissociative and (Non)-Completive Morphology in Totela Narrative
15
‘tomorrow I’m going [NONCMPL] to the fields’
b. ìjìlò ná-ndì-là-y-á kù-mpìlì
tom. POSTHOD-1SG-NONCMPL-go-FV CL17(LOC)-CL9.fields
‘tomorrow I’m going [POSTHOD] to the fields’
Consultants commented that speakers using forms such as (27b) “are still
hesitating” about whether they will go, suggesting less epistemic certainty
or strength of speaker expectation with na- forms.
It seems, then, that prehodiernal pasts without -ka- depict situations
whose completion (and the ensuing state of affairs) is relevant enough at
Perspective Time to be represented as within the domain of Perspective
Time; prehodiernal futures without na- have enough certainty at
Perspective Time to be seen as inevitable outcomes already in progress.
Both of these situations can reasonably be construed as within the domain
of Perspective Time.
Further evidence for the Associative/Dissociative domain distinction
may be seen in morphological similarities between markers of temporal
distance and markers of spatial distance or irreality. Prehodiernal -ka-
(28a) is segmentally and tonally identical to a distal -ka- marker (28b) in
Totela (although they are distinguished by different melodic tone patterns
in relative clauses). The two markers may also co-occur, as in (28c).
(28) a. ndà-kà-sàmb-à
1SG.CMPL-PREHOD-bathe-FV
‘I bathed (yesterday or before)’
b. ndà-kà-sàmb-à
1SG.CMPL-DIST-bathe-FV
‘I bathed (today, elsewhere from here)’
c. ndà-kà-kà-sàmb-à
1SG.CMPL-PREHOD-DIST-bathe-FV
‘I bathed (yesterday or before, elsewhere from here)’
Although the two -ka- markers likely had separate historical developments
(see Botne 1999 for discussion of similar markers across Bantu and Crane
2011 for more on ka markers in Totela), their morphological and semantic
similarity is striking. Both separate the situation content from the domain
of the utterance, whether temporally or spatially.
A somewhat more tenuous connection may be seen between
posthodiernal na- and a counterfactual (‘not real’) na- past prefix, as
shown in (29).
(29) kámbè bà-kèz-à,

Dissociative and (Non)-Completive Morphology in Totela Narrative
16
COUNTER 3PL-DIST.come-FV
ná-twà-tàb-á sùnù
COUNTER-1PL.CMPL-become.happy today
‘if they had come, we would have been happy today’
Dissociative ka can also be used in relative tense constructions to
establish a time frame within which Associative forms can be used to
indicate relative temporal distance from the context situation. An example
is given in (30). In (30a), where the relative event (a death) happened on
the same day as the framing event (an arrival), an Associative form is used
to mark the relative timing of the death. In (30b), in contrast, the death
occurred at least one day prior to the arrival, and a Dissociative form is
used with both verbs.
(30) a. Context: a person died on the morning of the speaker’s arrival
... ndà-kà-mù-wààn-à á-fw-à
1SG.CMPL-PREHOD-3SG.OM-find-FV 3SG.CMPL-die-FV
‘I found him dead’ (i.e. ‘I found him [having] died [CMPL]’)
b. Context: a person died on Tuesday; the speaker arrived on
Wednesday
... ndà-kà-wààn-à à-ká-fw-à kàléè
1SG.CMPL-PREHOD-find-FV 3SG.CMPL-PREHOD-die-FV already
‘I found that he already died [CMPL] (the day before)’
The use of -ka- shifts the Perspective Time to the speaker’s at the time of
arrival; further information is given from that Perspective Time and is
marked accordingly. It is argued below that Dissociative ka is used to set
up a narrative’s content as outside of the time, space, and (possibly) reality
of its telling. Within the narrative, Associative forms can be used to
indicate that the content occurs within the domain of the narrative world.
2.3 Summary of verb forms
This section has outlined the use of several verb forms in Totela,
arguing that -a- is a marker of nuclear completion, that -la- and forms
unmarked for TAM coincide with nuclear non-completion, and that ka and
na- are markers of temporal dissociation. The morphemes discussed are
repeated in Table 4, along with brief recapitulations of their interpretations
and analyses.
Table 4. Summary of forms discussed in §2
Form
Interpretations
Analysis

Dissociative and (Non)-Completive Morphology in Totela Narrative
17
-a-
perfective, perfect,
present stative
Situates Perspective Time after the
completion of the situation nucleus
-la- /
unmarked
present (and present
stative), progressive,
future
Situates Perspective Time before
the completion of the situation
nucleus
-ka- / ka-
prehodiernal perfective
/ imperfective
Marks the situation as excluded
from (and prior to) the temporal
domain of Perspective Time
na-
posthodiernal future
Marks the situation as excluded
from (and after) the temporal
domain of Perspective Time
I hope to have established in this section that category labels such as
‘perfective’ and ‘present’ cannot capture the uses of Totela tense/aspect
marking, or their interactions with various situation types. The above
analysis of these forms also better accounts for their uses in narrative.
Sections 3 and 4 describe how speakers use the -a-, -la-/unmarked, and ka
forms within narratives as tools for adding structure and provoking
audience response. I first turn to a discussion the occurrence of forms
marking nuclear (Non-)completion.
3 (Non-)completion in Totela narrative
Completive and Non-completive forms are found frequently throughout
Totela narratives, and play important roles that are extensions of their
basic temporal semantics. Section 3.1 discusses Completive marking and
its use in structuring the narrative. Section 3.2 shows how Non-
completives are used to pace the narrative and encourage listener empathy
and participation.
3.1 Completion in narrative
As argued in §2.1.2 above, the Completive -a- form indicates that a
situation’s main content (nucleus) has ended prior to Perspective Time.
The narrative functions of the marker, which follow from its Completive
semantics, make it a useful narrative-structuring device. A Completive
form in a narrative indicates that a situation, or set of situations, has
reached completion with respect to Perspective Time. Verbs with
Completive marking can thus be used to indicate episodic boundaries.
They are additionally used to mark resumptively-functioning predicates
(referring back to previous events or summarizing a set of events) as well
as verbs depicting situations whose completion is crucial to the ensuing
action.

Dissociative and (Non)-Completive Morphology in Totela Narrative
18
3.1.1 Completive marking as episodic boundary marking
Probably the most important role of Completive morphology in Totela
narratives is to structure the narrative as a set of semi-discrete episodes or
‘scenes’. The use of the Completive to partition narratives into episodes
may be seen in the micro-narrative in example (31), which begins a
description of a speaker’s activities and experiences of that day. The
speaker uses Completive forms to mark ‘boundaries’ between the mini-
scenes of the narrative.
(31) Sunu, twa-buuk-a! Twa-buuk-a
today 1PL.CMPL-wake.up-FV 1PL.CMPL-wake.up-FV
‘Today, we woke up [CMPL]! We woke up [CMPL],
nde-twa-y-a ku-ku-nyuk-a obwizu
DM-1PL.CMPL-go-FV CL17(LOC)-INF-cut-FV CL14.grass
that’s when we went [CMPL] to cut grass,
ku-ka-kamul-a bweñanda... 6
INF-DIST-cut(LOZ)-FV CL14(POSS).CL9.house
to cut grass for the roof.
Twa-ka-ßool-a [bwangu] ... ku-ka-bool-a
1PL.CMPL-DIST-return-FV [soon] NARR-DIST-return-FV
We came back soon... (When) we came back [NARR]
ku-ku-nyukul-a ku-iz-a ku-sik-a
CL17(LOC)-INF-cut-FV NARR-come-FV NARR-arrive-FV
from cutting grass, we came [NARR] and arrived [NARR]
ku-ba-waan-a abantu betu
NARR-CL2.OM-find-FV CL2.people CL2.1PL(POSS)
and found them [NARR], those people of ours,
abo ba-li-tu-libelele.
CL2.DEM CL2-PRES.STAT-1PL.OM-wait.for.STAT
they were waiting for us.
Nde-twa-kumbil-a echibaka kuti
DM-1PL.CMPL-ask.for-FV CL7.excuse that
That’s when we asked [CMPL] to be excused so
tu-chi-y-a ku-nkwaya.
1PL-PERS-go-FV CL15(LOC)-CL9.choir
that we could first go to choir practice [before doing research
activities].’
The main topics dealt with in this brief narrative include going to cut
grass, returning to the village and what was found there, and requesting a
change of plans from the researcher. The speaker uses a Completive form
with each shift in topic or location, which marks the beginning of a new
6 Ellipses indicate utterances that are not reproduced here. They usually involve
repetition of the same information, with the same or similar verbal marking.

Dissociative and (Non)-Completive Morphology in Totela Narrative
19
mini-episode.
The passage begins with a Completive predicate, twabuuka ‘we woke
up’, which opens the narrative and, together with the adverb sunu ‘today’,
sets the main narrative time frame (see more on setting the time frame of a
narrative in §4 below). The speaker then immediately uses a Completive
form (ndetwaya... ‘that’s when we went...’) to introduce the first episode,
going to cut grass. After describing the grass-cutting activity, she uses
another Completive form, twakaboola ‘we returned from there’ (this -ka-
is a distal form), to indicate another shift between activities and locations.
This form is followed by a series of narrative forms, all describing the
events that occurred upon arriving at the village, as well as a stative form
(balitulibelelele ‘they’re waiting for us’) that describes a situation that was
happening in the background of the speaker’s own progression of actions.
The excerpt is brought to a close with a Completive utterance
(ndetwakumbila ‘right then we asked’), indicating a shift from planned
and expected activities (working with the researcher) to a new set of
activities (choir practice).
The Completive’s use in marking episode boundaries is seen
especially in its very common co-occurrence with verbs of coming and
going, such as -ya ‘go’, -boola ‘return’, and -sika ‘arrive’. These types of
verbs are used to shift, via character movement, the story’s action from
one location to another. They often coincide with episode boundaries. For
example, (24) shows the ending of a narrative scene. In this scene, an old
hag tricks a young mother into handing over her baby.
(32) “Àngù ndì-là-bòòl-à.” Ii!
soon 1SG-NONCMPL-return.FV INTERJ
‘“I’ll come right back [NONCMPL].” And she [the young woman]
Nòkú-sùm-ùnùn-à kù-mú-p-à.
COM.NARR-tie-REVERS-FV NARR-3SG-give-FV
untied [NARR] [the baby from her back] and gave it to her [NARR].
Á-y-à.
3SG.CMPL-go-FV
She [the old hag] left [CMPL].’ (Kalima Mawundu)
The narration following this excerpt describes the parallel activities of
the hag and the baby in one location, and the mother in another.
Completive àyà ‘she left’ marks the boundary. In this case, the use of
Completive marking also marks a change in participants from the young
woman to the old hag, but participants also frequently change within a
string of forms with narrative marking (participants are often, though not
universally, overtly named in such cases—see, for example, (35) below),
and the Completive can also be used without a change in participants.

Dissociative and (Non)-Completive Morphology in Totela Narrative
20
The use of the Completive with verbs of coming and going is a
consequence of its use as a boundary marker, and not merely an
epiphenomenon of location-changing predicates being used as episode
boundaries. This can be seen in the following examples, in which the
Completive is used in boundary predicates that are not location changing
(33–34), and in location changing predicates that do not mark episode
boundaries (35).
Examples (33) and (34) are from the story of an animal skin that
magically transforms into a woman and does housework every day while
the owner is away. Each day’s work comes to an end when the owners
arrive and the woman has to transform back into an animal skin to protect
her secret identity; the arrival is marked with a Completive form (bàsìkà
‘they arrived’).
(33) Kù-màn-à kù-zàbik-à pèlè nò-ká-kòlàm-à
NARR-finish-FV INF-soak-FV PART COM[NARR]-DIST-ascend-FV
‘She finished [NARR] soaking [the millet], then went up [NARR]
kóòkó èkál-à, kù-màlìbélà
CL17.DEM 3SG.stay-FV CL17(LOC)-CL6.corner
to where she stays [NONCMPL.RC], in the corner
éèñándà. Nàbó... àbénì...
CL6(POSS).CL9.house COM.CL2.DEM CL2.owner
of the house [stuck in the roof]. And those owners
béèñándà bà-sìk-à.
CL2(POSS).CL9.house CL2.CMPL-arrive-FV
of the house arrived [CMPL].’ (Kanyama)
(34) Aa! Nàbó àbénì bà-sìk-à.
INTERJ COM.CL2.DEM CL2.owner CL2.CMPL-arrive-FV
‘Aa! Then those owners arrived [CMPL].
Nàkó nòkú-yàngàm-á mwìjúlù.
COM.CL12.DEM COM.NARR-raise.(self).up-FV L18(LOC).CL5.top
And it [the little skin/woman] hung herself up [NARR] in the
roof beams.’ (Kanyama)
Example (35) comes from the final scene of a narrative, after villagers
arrive in a field and discover the aftermath of a crime and run back to call
the others to come look. The journey to and from the village does not
constitute an episode change, and is marked with narrative, rather than
Completive, morphology.
(35) Nòkù-làwùk-à ò-ká-kùw-á àbàntù
COM.NARR-run-FV COM.NARR-DIST-call-FV CL2.people

Dissociative and (Non)-Completive Morphology in Totela Narrative
21
‘Then they ran [NARR] and called [NARR] the people
kù-mùnzì ... Nò-kèz-à kwìz-à
CL17(LOC)-CL3.village COM.NARR-DIST.come-FV NARR.come-FV
in the village... They [the villagers] came [NARR]
kù-wààn-à...
NARR-find-FV
and found [NARR]...’ (Fumako)
In addition to its occurrences at scene boundaries, the Completive’s
non-occurrence within scenes provides further evidence of its episode
boundary marking function. Of particular interest is its non-occurrence
following songs.
Songs are a vital part of storytelling across Africa (see e.g. Finnegan
2013: 419ff, who notes that “without [songs], in many cases, the stories
would only be a bare framework of words” (419)). When performing a
Totela folk narrative, speakers frequently employ songs to develop both
the storyline and the characterization of participants, as well as to invite
listeners to actively join the storytelling process. Characters often have
‘signature’ songs, which may change somewhat in content as the story
moves forward. Songs also set the mood, foreshadow coming events, and
enliven the action. Typically, a speaker will start a song and sing any
changing verses, with the audience joining in on choruses.
Even as songs move the storyline forward, they are somewhat
disruptive to story action. However, they typically occur within an
episode, rather than at its boundary. Correspondingly, the Completive is
used quite rarely with verbs immediately following songs in narratives.
Instead, consecutive narrative forms are almost universally used, as shown
in example (36), taken from the comically gruesome story of a father who
murders his child and is chased by the child’s singing disembodied head.7
(36) Sà-ù-y-á bù-yîmb-à
DM-CL3-go-FV CL14-sing-FV
‘It [the child’s head] went along [NONCMPL] singing [MANNER].
/ [SONG] /
Aa! Nòkù-ìnd-á èchíkàtà kù-sòw-à
INTERJ COM.NARR-take-FV CL7.carrier NARR-throw.away-FV
Aa! Then [the father] took [NARR] the carrying wrap [from his
head] and threw [it] away [NARR].’ (Namayoyo)
7 As also shown in this example, songs are almost always introduced with a Non-
completive form. This use is discussed in §3.2.1.

Dissociative and (Non)-Completive Morphology in Totela Narrative
22
In calling the audience’s attention back to the episode-in-progress, the use
of a narrative form, rather than a boundary marking Completive, helps to
increase within-scene continuity (cf. Nurse 2008: 120, who notes that
narrative forms can be used to “stress continuity”).
3.1.2 Resumptive and other uses of Completive marking
As an indicator of episode boundaries, the Completive functions on a
textual, information-structural level. The Completive also occurs in
clauses where its use is both textual and linked to narrative-internal time,
either marking non-sequential events—i.e., clauses in which narrative
morphology would be inappropriate—or functioning resumptively,
recalling or summarizing important events already described in the
narrative. While such uses are not always optionally replaced by narrative
morphology, a speaker’s choice to use a resumptive clause is nonetheless
significant in structuring the narrative, especially in calling (or recalling)
the listeners’ attention to important points in the story.8
Resumptive use of the Completive is especially common with the
verb -mana ‘finish’ in temporal clauses or summaries of a state of affairs.
These uses are illustrated in (37) and (38), where they signal the
completion of something that will be important for the rest of the story.
For example, the birth in (37) and the mother’s subsequent actions set the
stage for the abduction witnessed in example (32) above. Use of the
Completive and the repeated naming of the situation mark the birth as a
significant action within a sea of narrative predicates, setting in motion all
the events to come.
(37) Áàwò kà-lì mùntù. Ii!
CL16(LOC).DEM IPFV.CL1-be CL1.person INTERJ
‘Once, there was [PREHOD.IPFV] a man. Ii!
Nòkú-sès-á òmwánàkázì,
COM.NARR-marry-FV CL1.woman
And he married [NARR] a woman,
kwíbbàl-ìl-a=kó òmwáànà. Ii!
NARR.bear-APPL-FV=CL17(LOC) CL1.child INTERJ
and she bore him [NARR] a child. Ii!
À-màn-à kwíbbàl-á òmwáànà...
CL1.CMPL-finish-FV INF.bear-FV CL1.child
When she had finished [CMPL] giving birth to the child...’
(Kalima Mawundu)
In (38), which is drawn from a description of daily routines in past
times, use of the Completive with -mana similarly marks the pounding of
8 In contrast to narrative forms being used after songs to stress narrative continuity,
‘resumptive’ clauses (used in the sense of e.g. Seidel 2008: 380) provide overall structure
and portray situations as particularly important.

Dissociative and (Non)-Completive Morphology in Totela Narrative
23
grain into flour as a necessary predecessor to what follows.
(38) Esi twa-man-a oku-twa ku-ijik-a
COND 1PL.CMPL-finish-FV INF-pound.FV NARR-cook-FV
‘When we (had) finished [CMPL] pounding, we cooked [NARR]
ku-li-a.
NARR-eat-FV
and ate [NARR]’ (Akale-kale)
Example (39) is taken from a procedural description of daily work. The
Completive is used several times, referring back to previously completed
actions that were necessary for what followed. Note that -mana can also
be used non-resumptively, and indeed is used with narrative forms
throughout the example.9
(39) Ku-man-a ku-sanz-a. Esi
NARR-finish-FV NARR-bathe.CAUS-FV COND
‘We finish [NARR] washing [NARR] [the millet]. When
a-bomb-a noku-leet-a ku-twa.
CL6.CMPL-soak-FV COM.NARR-bring-FV NARR-pound.FV
it’s soaked [CMPL] then we take [NARR] it and pound [NARR] [it].
Ku-man-a ku-twa ku-leet-a
NARR-finish-FV NARR-pound-FV NARR-bring-FV
We finish [NARR] pounding [NARR] and bring [NARR] [it]
kwijik-a insima.
NARR.cook-FV CL9.porridge
and cook [NARR] nshima.
Wobulya obusu twa-twa
CL16(LOC).CL14.DEM CL14.flour 1PL.CMPL-pound.FV
That flour we pounded [CMPL],
ku-tuul-a mu-mpoto kwijik-a.
NARR-put-FV CL18(LOC)-pot NARR.cook-FV
we put [NARR] [it] in the pot and cook [NARR] [it].
Ku-man-a kwijik-a ku-tuul-a
NARR-finish-FV NARR-cook-FV NARR-put-FV
(When) we finish [NARR] cooking, we put [NARR] [it]
a-misuba. Alya a-misuba
CL16(LOC)-CL9.pot CL16(LOC).DEM CL16(LOC)-CL9.pot
9 Note that the perfect/present stative use of Completive -a- can also be seen in this
example (esi abomba ‘when it’s soaked’).

Dissociative and (Non)-Completive Morphology in Totela Narrative
24
on plates. Those plates
twa-tuul-a nelo nde-tu-ind-a ku-li-a.
1PL.CMPL-put-FV now DM-1PL-take-FV NARR-eat-FV
we put [CMPL] [it on], that’s when we take [them]
[NONCMPL/UNMARKED] and eat [NARR]’ (Emisebezi yetu)
3.1.3 Completion in Totela narrative: summary
The semantics of the Completive form in Totela make it a valuable
discourse-structuring device. Indicating that a situation—or, by extension,
a related set of situations—has reached completion with respect to
Perspective Time, it highlights the importance of that completion. It marks
an episode boundary or signals that a completed situation is important for
what follows.
3.2 Non-completion in narrative
Non-Completive forms, whether morphologically unmarked or marked
with ‘disjunctive’ -la-, also play an important role in structuring Totela
narratives. Rather than being linked to narrative-internal time, Non-
completive forms serve to invite empathy and response from the listeners,
as if temporarily pulling the narrative content from the distant time and
place of its setting (see §4) to the here-and-now of the storytelling setting.
3.2.1 Non-completion and songs
As noted in section 3.1.1, songs form an integral part of many Totela
folk tales, and are often repeated throughout the narrative, sometimes with
subtle changes reflecting the current state of affairs. They depict character
personalities and desire, and some narratives have different songs for
different situations. They can be the means for invoking magical powers to
provide solutions. They also invite audience participation, as listeners
typically join in after the first chorus.
Nearly every song in my corpus of Totela narratives is introduced
with a form involving the discourse-marking prefix se- (or sa-) and a Non-
completive predicate. Speakers translate se- as indicating ‘the very next
thing’ or ‘what followed after’ or ‘then X began...’. Examples (40) and
(41) are illustrative of typical progressions leading up to songs.
(40) Sà-lì-kà-ìlè. Nàwó òmútwì
DM-PRES.STAT-DIST-go.STAT COM.CL3.DEM CL3.head
‘There he goes [STAT]. And that head [following him]
sà-ù-lá-yìmb-à: / [SONG]
DM-CL3-NONCMPL-sing-FV

Dissociative and (Non)-Completive Morphology in Totela Narrative
25
sings [NONCMPL]:’ [SONG] (Fumako)
(41) Kù-kà-sìk-à=kóò
NARR-DIST-arrive-FV=CL17(LOC)
‘They [the children] got there [to the tree] [NARR]
ò-kà-tànt-à ò-kà-bùlùmùn-à
COM.NARR-DIST-climb-FV COM.NARR-DIST-knock.down-FV
and climbed up [NARR] and [shook it and] knocked down [NARR]
ò-kà-bùlùmùn-à.
COM.NARR-DIST-knock.down-FV
and knocked down [NARR] [the fruit].
Ò-kézùzy-à [ezinkalanga].
COM.NARR-DIST.become.full.CAUS-FV [CL8.basket]
They filled [NARR] [their baskets].
Nàlyó èkìshíkìshì nòkù-sìk-à.
COM.CL5.DEM CL5.goblin COM.NARR-arrive-FV
And that ekishikishi also arrived [NARR].
Sà-bà-lá-yìmb-à: / [SONG]
DM-3PL-NONCMPL-sing-FV
Then the children begin to sing [NONCMPL]: / [SONG]
(Kumangwembya)
The Non-completive (along with se-/sa-) in these cases performs
several functions. First, se- marks the song as a response to the previous
narrative events. The Non-completive form indicates an incomplete
situation: the song represents a break in the actions currently underway.
Furthermore, the shift from a narrative form to an inflected, Non-
completive form temporarily links the time frame of the narrative to the
time of the telling, inviting the listeners to participate.
Other examples of the Non-completive in narrative temporarily slow
the narrative pace and invite listener empathy, portraying situations and
events more vividly than simply another narrative form in a string of
narrative forms would portray them. For example, (42) comes after a man
has just married and brought his wife home.
(42) ...òkù-mú-twààl-á kù-mpìlì.
INF-3SG.OM-carry-FV CL17(LOC)-CL9.field
‘...he took her to the fields.
À-lá-kàngwà.
3SG-NONCMPL-be.impossible-PASS-FV

Dissociative and (Non)-Completive Morphology in Totela Narrative
26
She can’t do it [farming work] [NONCMPL].’ (Lobwa Lobwa)
The shift to a Non-completive form seems to invite the listener to take
the perspective of the husband, and to experience his dismay. In the
recording of example (42), the utterance àlákàngwà (‘she fails’/ ‘she can’t
do it’) is preceded by a brief pause, and followed by a pause of nearly two
seconds, furthering the sense that this is a statement of import. Indeed, the
failure of the wife as a farm worker is an ominous sign: she is later
discovered to be a zebra.
Another example of the use of the Non-completive as a pacing device
and to increase vividness comes from the story, already mentioned in
examples (36) and (40), of a disembodied head following its murderous
father. Having thrown away all of his possessions in an effort to get away
more quickly, the desperate father begins to cut off his own limbs during
the chase.
(43) Aa! Nòkwìnd-á èchìbèlò kù-kòsòl-à
INTERJ COM.NARR.take.FV CL7.leg NARR-cut-FV
‘Aa! Then he took his leg [NARR] and cut it off [NARR]
kù-sòw-à ... Sà-yènzy-á ìtèndè lyònkéè.
NARR-throw-FV DM.3SG-walk.CAUS-FV CL5.leg CL5.one
and threw it aside [NARR]. Then he’s walking [NONCMPL] on just
one leg.’ (Fumako)
In summary, use of the Non-completive in Totela narrative allows the
speaker to temporarily slow the pace of the narrative and to encourage
listener empathy and participation by aligning narrative-internal time with
the time of the storytelling. While the Completive marks the boundary of
an episode (‘this is finished’), the Non-completive not only marks
something as not finished, but also brings it into the here-and-now of the
narrative’s audience.
4 Dissociation in narrative
While the Completive and Non-completive appear with some
frequency throughout narratives, and are used to structure and pace them,
Dissociative forms are found only at the beginnings and ends of
narratives—virtually never in the main body—and serve to situate the
narrative in (possibly fictional) time and space. This section deals with
Dissociative prehodiernal ka; posthodiernal na- does not appear in my
narrative corpus outside of the quoted speech of characters.

Dissociative and (Non)-Completive Morphology in Totela Narrative
27
4.1 Abstract and orientation
In the corpus of narratives, the first verb is almost universally a
Dissociative (prehodiernal imperfective or Completive) form, followed
either by a series of such forms, or by a narrative form. Examples (44–46)
show the opening lines of three narratives. Each starts with a formulaic
‘once there was [PREHOD.IPFV]’. In (44) and (45), several more verbs
marked with a prehodiernal form set the stage, giving background
information on the characters and their circumstances. Example (46)
plunges straight into the narrative action after introducing one of the
characters.
(44) Áàwò kà-lì mùchèchè
CL16(LOC) PREHOD.IPFV-be CL1.child
‘There once was [PREHOD.IPFV] a child,
à-ká-fwìlw-á bà-nyìnà.
CL1.CMPL-PREHOD-die.APPL.PASS-FV CL2A-his.mother
his mother died on him [PREHOD.CMPL].
Éènì chwàlè nìngá bà-ká-fwà
yes INTERJ NINGA CL2A.CMPL-PREHOD-die.FV
Yes. Well, when she died [PREHOD.CMPL],
bà-nyìnà, ná-à-ká-sìyàl-á
CL2A-his.mother DM-CL1.CMPL-PREHOD-be.left-FV
his mother, that’s when he was left [PREHOD.CMPL]
kù-bésì. Bési nokú-ses-a
CL17(LOC)-CL2A.his.father CL2A.his.father COM.NARR-marry-FV
with his father. Then his father married [NARR]
yu-mwí omwánakázi.
CL1-other CL1.woman
another woman.’ (Kañandu)
(45) Kà-tù-lì tùchèchè nàkàchèmbèlè
PREHOD.IPFV-CL13-be CL13.child COM.CL12.old.person
‘There once were [PREHOD.IPFV] children, and a little old lady
ká-kèèkál-à mu-chiole
PREHOD.IPFV-CL12.stay-FV.RC CL18(LOC)-CL7.bush
who lived [PREHOD.IPFV] in the bush,
mu-mutemwa. Eènì.
CL18(LOC)-CL3.forest yes
in the forest. Yes.
Kà-ká-pòn-áàng-á òbwíìlì. ...
PREHOD.IPFV-CL12-live-HAB-FV.RC CL14.tuber

Dissociative and (Non)-Completive Morphology in Totela Narrative
28
She lived [PREHOD.IPFV] on obwiili roots.
kà-ká-yààk-à nèñándà. Íngà
CL12.CMPL-PREHOD-build-FV COM.CL9.house DM
And she also built [PREHOD.CMPL] a house.
nàtó !tóòtò òtùchèchè
COM.CL13.DEM CL13.DEM CL13.child
And so those children
kà-tù-yèmbèl-áàng-à mù-mútèmwà
PREHOD.IPFV-CL13-herd-HAB-FV.RC CL18(LOC)-CL3.forest
who herded [PREHOD.IPFV] in the forest
nòkwíjàlwil-á òtùpóngò.
COM.NARR-open.APPL-FV CL14.goats
let out [NARR] the goats. (Namayoyo)
(46) Áàwò kà-lì mú!kwáámè
CL16(LOC).DEM PREHOD.IPFV.CL1-be CL1.man
‘There once was [PREHOD.IPFV] a man,
nò-kà-tòbèl-á òmwánàkázì. Kù-mú-lèèt-á
COM.NARR-DIST-seek-FV CL1.woman NARR-CL1.OM-bring-FV’
he went to seek [NARR] a wife. He brought her [NARR]
à-mùnzì.
CL15(LOC)-CL3.village
to the village.’ (Choncho)
In the above examples, imperfective ka- forms give information about
the setting and the state and habits of the participants. The Completive
-a-ka- forms detail actions leading up to the story line. Together, this
information forms what Labov & Waletzky (1967) call the story’s
‘orientation’. Orientations in Totela narratives not only give background
information, but the use of Dissociative forms in them also signals to
listeners that the story that is about to be told takes place in a time and
place—and, in the case of fictional folk narratives, another reality—than
the time of its telling.
While the stories in the narrative corpus generally open with an
orientation similar to those in examples (44–46), the orientation section is
occasionally preceded by what Labov & Waletzky (1967) term ‘abstract’,
that is, a brief initial summary or introduction. Narrative abstracts, when
they occur, also use prehodiernal Dissociative forms, as in example (47).
(47) Mwáà-ká-chìt-íl-à
CL18(LOC).CL1.CMPL-PREHOD-do-APPL-FV.RC

Dissociative and (Non)-Completive Morphology in Totela Narrative
29
‘This is how he acted [PREHOD.CMPL],
yóòyó òmùntù …
CL1.DEM CL1.PERSON
that person...’ (Nyanyambe)
Example (47) is followed by a short orientation, and then the main action
commences.
As noted above, after the orientation, Dissociative forms virtually
never appear within the main narrative body outside of quoted speech.
Instead, the narrative and inflected forms detailed in previous sections are
used. The orientation sets the scene, and then the imaginary world of the
narrative becomes the base world from which Perspective Times are set.
4.2 Coda
At the end of almost every narrative, the Dissociative appears once
again, this time in a conventionalized ‘coda’, as in (48).
(48) Pó-lwà-kà-màn-ín-à
CL15(LOC)-CL11.CMPL-PREHOD-finish-APPL.FV
‘That’s where the story ended’ (Kanyama)
Labov (2001: 65) describes the function of a coda as indicating
“termination of the narrative by returning the time frame to the present”.
Labov & Waletzky note that this effect is often achieved through use of
deixis, for example, “obviate” ‘this’ and ‘those’ in contrast to the
“proximate” ‘these’ and ‘here’ that are used throughout the man narrative
(Labov & Waletzky 1967: 100). Dissociative marking on the verb fits this
description neatly. It shifts Perspective Time from the story time set by the
orientation, back to the here-and-now of telling. It reminds the listeners
that the narrated events are removed from their own time, space, and
reality.
4.3 Dissociation in narrative: summary
During a narrative, listeners suspend disbelief and are figuratively
transported to the cognitive domain of the narrative. The use of
Dissociative markers in the narrative’s abstract and orientation signal this
shift. At the end of the narrative, another Dissociative marker in the coda
closes the domain of “story-reality” and brings listeners back to the

Dissociative and (Non)-Completive Morphology in Totela Narrative
30
present world of the storytelling. In framing a narrative, Dissociative
marking delineates a world, separate from the world of telling, where the
narrative events—fantastic or mundane—take place. Inside that world,
associative forms reflect story-internal reality.
5 Conclusion
Table 5 summarizes the TAM categories discussed in this article and
their specific uses in narrative, as shown in the preceding sections.
Table 5. TAM categories and their uses in narratives
Category
General Use
Narrative Use
DISSOCIATIVE
(PREHODIERNAL)
Mark relevant time of
situation and its aftermath
as prior to the day of
utterance time
Situate a narrative as
elsewhere in time, space,
and reality status; shift from
narrative-internal time back
to time of storytelling
COMPLETIVE
Situate Perspective Time
after nuclear completion
Mark episode boundaries;
mark completion of events
that will be significant in
the development of the
narrative; resumptive uses
NON-COMPLETIVE
Situate Perspective Time
as prior to nuclear
completion
Slow the pace of the
narrative; invite listener
empathy and participation
by aligning listener
perspective and narrative-
internal perspective
NARRATIVE
MORPHOLOGY
Indicate consecutive
actions
Indicate consecutive
actions; increase sense of
continuity within episodes
Comparison of the “general uses” and “narrative uses” in each row of
Table 5 makes it clear that the narrative uses are closely related to the
forms’ general uses, and indeed may paint a fuller picture of the true
nature of the categories. Language is not a series of single utterances
isolated in a vacuum, but rather a living vehicle of narration and
conversation. The narrative uses described in this paper provide further
evidence for the need to understand tense and aspect within systems such
as the approach proposed by Botne & Kershner (2008). The Totela data
show that it is enlightening—and, in fact, necessary—to view tense and
aspect marking holistically, taking into account all of their contexts, rather
than as purely temporal phenomena.
Abbreviations

31
1/2/3 1st / 2nd / 3rd person
APPL applicative
C coda of situation
CAUS causative
CL1 noun class 1
CL1A noun class 1a
CMPL completive
COM comitative (‘with’/‘and’)
COND conditional
COUNTER counterfactual
DIST distal
DM discourse Marker
FV final vowel
HAB habitual
INF infinitive
INTERJ interjection
IPFV imperfective
LOC locative
N nucleus of situation
NARR narrative morphology
NONCMPL non-completive
O onset of situation
OM object marker
PASS passive
PO pratical orthography
PRON pronoun
PL plural
PT perspective time
RC relative clause
REVERS reversive
SG singular
SM subject marker
STAT stative/stativizer
References
Beuchat, Phyllis-Doris. 1966. The Verb in Zulu. Johannesburg:
Witwatersrand University Press.
Botne, Robert. 1983. On the notion ‘inchoative verb’ in Kinyarwanda. In
Le Kinyarwanda, langue bantu du Rwanda: Etudes linguistiques, F.
Jouannet (ed), 149-180. Paris: Société d'Études Linguistiques et
Anthropologiques de France.
Botne, Robert. 1999. Future and distal -ka-’s: Proto-Bantu or nascent
form(s?). In Bantu Historical Linguistics: Theoretical and Empirical
Perspectives [CSLI Lecture Notes], Jean-Marie Hombert & Larry M.
Hyman (eds), 473-515. Stanford, CA: CSLI Publications.
Botne, Robert. 2010. Perfectives and perfects and pasts, oh my!: On the
semantics of -ILE in Bantu. Africana Linguistica 16: 31-63.
Botne, Robert. 2012. Remoteness distinctions. In The Oxford Handbook of
Tense and Aspect, Robert Binnick (ed), 536–562. Oxford: Oxford
University Press.
Botne, Robert & Kershner, Tiffany L. 2000. Time, tense, and the perfect
in Zulu. Afrika und Übersee 83: 161-180.
Botne, Robert & Kershner, Tiffany L. 2008. Tense and cognitive systems:
on the organization of tense/aspect systems in Bantu languages and
beyond. Cognitive Linguistics 19: 145-218.
Carlson, Robert. 1992. Narrative, subjunctive, and finiteness. Journal of
African Languages and Linguistics 13: 59-85.
Condoravdi, Cleo. 2002. Temporal interpretation of modals: Modals for
the present and for the past. In The construction of meaning, David

Dissociative and (Non)-Completive Morphology in Totela Narrative
32
Beaver, Luis D. Casillas Martínez, Brady Z. Clark, & Stefan
Kaufmann (eds), 59-87. Stanford: CSLI Publications.
Cover, Rebecca. 2010. Aspect, Modality, and Tense in Badiaranke. PhD
dissertation, University of California, Berkeley.
Crane, Thera. 2011a. Beyond Time: Temporal and Extra-temporal
Functions of Tense and Aspect Marking in Totela, a Bantu Language
of Zambia. PhD dissertation, University of California, Berkeley.
Crane, Thera. 2011b. “Completion” semantics in Totela tense and aspect.
Paper presented at the Fourth International Conference on Bantu
Languages, Berlin.
Crane, Thera. 2011c. Completion and dissociation in Totela tense and
aspect. In Proceedings of the 42nd Annual Conference on African
Linguistics, Michael R. Marlo, et al. (eds), 208-220. Somerville, MA:
Cascadilla Proceedings Project.
Crane, Thera. 2013. Resultatives, progressives, statives, and relevance: the
temporal pragmatics of the -ite Suffix in Totela. Lingua 133: 164-188.
Dahl, Östen. 1985. Tense and Aspect Systems. Oxford: Blackwell.
Dahl, Östen. 2008. The morning is wiser than the evening. Paper
presented at CHRONOS 8: International Conference on Tense, Aspect,
Mood, and Modality, Austin, Texas.
Dahl, Östen. 2009. ‘Today’ and ‘yesterday’ in tense systems. Paper
presented at the Eight Biennial Conference of the Association for
Linguistic Typology, Berkeley, California.
Finnegan, Ruth. 2013. Oral Africa in Literature. Cambridge: Open Book
Publishers. < https://play.google.com/store/books/details?id=
T4ZPAgAAQBAJ&rdid=book-T4ZPAgAAQBAJ&rdot > (Free
download from Google Play, 29 January 2015. Revised version of
Finnegan 1970.)
Fleischman, Suzanne. 1990. Tense and Narrativity: From Medieval
Performance to Modern Fiction [Texas Linguistics Series]. Austin:
University of Texas Press.
Fortune, G. 1949. The conjugation of inchoative verbs in Shona. African
Studies 8: 132-140.
Freed, Alice. 1980. The Semantics of English Aspectual Complementation.
Dordrecht: D. Reidel.
Hatav, Galia. 2012. Bound tenses. In The Oxford Handbook of Tense and
Aspect, Robert Binnick (ed), 611-637. Oxford: Oxford University
Press.
Hopper, Paul J. 1979a. Aspect and foregrounding in discourse. In
Discourse and Syntax, Talmy Givón (ed), 213-241. New York:
Academic Press.
Hopper, Paul J. 1979b. Some observations on the typology of focus and
aspect in narrative language. Studies in Language 3: 67-64.
Hopper, Paul J. 1982. Aspect between discourse and grammar: an

Dissociative and (Non)-Completive Morphology in Totela Narrative
33
introductory essay for the volume. In Tense-Aspect: Between
Semantics & Pragmatics (Containing the Contributions to a
Symposium on Tense and Aspect, held at UCLA, May 1979, Paul J.
Hopper (ed), 3-18. Philadelphia: John Benjamins.
Hutchinson, Larry G. 1985. Nonlinguistic beliefs in linguistics. Journal of
Pragmatics 9: 109-136.
Labov, William & Waletzky, Joshua. 1967. Narrative analysis: oral
versions of personal experience. In Essays on the Verbal and Visual
Arts, June Helm (ed), 12-44. Seattle: University of Washington Press.
Langacker, Ronald W. 2000. Grammar and Conceptualization.
Berlin/New York: Mouton de Gruyter.
Nurse, Derek. 2008. Tense and Aspect in Bantu. New York: Oxford
University Press.
Reichenbach, Hans. 1947. Elements of Symbolic Logic. New York: The
Free Press.
Steele, Susan. 1975. Past and irrealis: just what does it all mean?
International Journal of American Linguistics 41: 200-217.
Sweetser, Eve & Fauconnier, Gilles. 1996. Cognitive links and domains:
basic aspects of mental space theory. In Spaces, Worlds, and
Grammar, Gilles Fauconnier & Eve Sweetser (eds), 1-28. Chicago:
University of Chicago Press.
Vendler, Zeno. 1957. Verbs and times. The Philosophical Review 66: 143-
160.