Genre, verb, and coherence in picture-elicited discourse of adults with aphasia

Abstract

Background:

Discourse coherence is derived, in part, from the relationship between and among words and sentences. In studies of aphasia, the relationship between discourse-level and sentence-level phenomena may be examined through the verb. In clinical picture elicitations of discourse, the nature of the pictures or the accompanying elicitation instructions may influence the discourse genre of the response (descriptive vs narrative), which in turn may place different linguistic demands (e.g., verb production demands) on the speaker with aphasia.

Aims:

This study explores aphasic speakers’ partial construction of discourse coherence through relationships between and among action, background, and setting information carried largely by the verb, in picture-elicited discourse productions of two different genres: descriptive and narrative.

Methods & Procedures:

Twelve individuals with aphasia of mild to moderate severity were presented with five pictures: two composite and three complex. Verbal discourse responses were elicited with common clinical instructions. For the three complex pictures, discourse was also elicited with instructions that explicitly requested temporal sequencing. Discourse genre (descriptive vs narrative) of each response was determined. Verbs within each response were categorised on dimensions of form (tense) and function (degree of association with storyline).

Outcomes & Results:

For common clinical elicitations, discourse genre was descriptive, verb tense was present and/or non-finite, and verbs filled background and setting functions, regardless of picture type. For elicitations that requested temporal sequencing, participants produced more narrative discourse genre, and more past tense and storyline function on verbs. Exceptions to the group patterns are discussed and exemplified.

Conclusions:

The patterns of group results and the exceptions to these patterns provide insights into the relationship between the discourse and sentence levels, and the ability of individuals with aphasia to negotiate this relationship in their discourse productions. Findings also hold methodological implications for sampling discourse production among individuals with aphasia.

Discourse is a language unit whose organisation supersedes any single word or sentence. It is the relationship between and among words and sentences, among other factors, which contributes towards the organisation of discourse. Discourse organisation is realised as a global coherence or “hanging together” of the text as a unit. The current study explores one potential contribution to discourse coherence, namely the relationships between and among action, background, and setting information carried largely by the verb. It examines the nature of these relationships in picture-elicited discourse productions of individuals with aphasia.

Aphasia research has historically neglected the relationship between sentence-level and discourse-level phenomena (Armstrong, 2000; Ulatowska & Olness, 2000). The verb/clause may be useful for examining this relationship. Verbs are recognised to play a key role at the level of the sentence, and at the level of discourse organisation (Longacre, 1996). For individuals with aphasia, studies suggest that the verb may be compromised at the word and sentence level (e.g., Bastiaanse & Edwards, 2004; Bastiaanse & Zonneveld, 1998). There is also evidence to suggest that the temporal-causal event line of narrative discourse, carried primarily by the verb, may be abbreviated in aphasia, at least for personal narratives on emotive topics (Ulatowska, Olness, & Williams, 2004). To complement this research, the current study examines the core temporal-causal event line of narrative, as well as other functions filled by the verb in discourse, in picture-elicited contexts.

Pictorial stimuli have been a traditional means of clinical discourse elicitation (Bottenberg, Lemme, & Hedberg, 1987; Myers & Linebaugh, 1984; Nicholas & Brookshire, 1993; Potechin, Nicholas, & Brookshire, 1987; Ulatowska & Chapman, 1991). Previous research indicates that single pictures, in particular, are likely to elicit a descriptive discourse genre, in which actors and actions are enumerated with little or no connection between them (Bottenberg et al., 1987; Olness, Ulatowska, Wertz, Thompson, & Auther, 2002). Armstrong (2000) calls specifically for greater attention to discourse genres and the language requirements of these genres, and examination of the ability of speakers with aphasia to meet these requirements. The current study examines the linguistic demands of the descriptive discourse genre elicited by single pictures. It also explores the degree to which single pictures can elicit a narrative genre, which has a temporal-causal organisation, thus changing the linguistic demands on the speaker, through the use of pictures and elicitation instructions that may foster a response with temporal-causal organisation.

With regard to selection of pictures for clinical discourse elicitation, Luria (1962/1966) highlights a differentiation between picture types that will be termed in the current study “composite” pictures and “complex” pictures. Composite pictures directly depict all the details about actors and actions that have little or no relationship to each other. One would expect this type of picture to elicit descriptive discourse, i.e., enumeration of actors and actions with little or no expression of the relationships between and among them. In contrast, complex pictures are those whose meaning can be interpreted only through a synthesis of the relationship between and among actors and their actions. Complex pictures might be expected to elicit more than simple enumeration in discourse responses. In particular, certain complex pictures may depict a complication or climax that represents the peak of action, or what photographers refer to as the “decisive moment” (Cartier-Bresson, 1952). From such pictures, the viewer may be able to infer what happened prior to the climax, and what may happen after the climax (Nicholas & Brookshire, 1993). This inferred temporal-causal progression can form the basis of a discourse response in the narrative genre, the core of which has a temporal-causal organisation.

With regard to discourse elicitation instructions, typical clinical instructions with single pictures (e.g., “Tell me everything you see going on in this picture”) do not make an explicit request for a temporal organisation. This makes a discourse response in the narrative genre unlikely. The current study explores whether instructions that explicitly request a temporal-causal line may result in responses in the narrative genre.

This study was designed to explore potential contributions to global coherence in picture-elicited discourse produced by individuals with aphasia, by posing the following questions:

1. Is the genre of discourse produced by individuals with aphasia in response to single-picture stimuli different for:

   1. pictures that depict a complication or climax, and those that do not; and
   2. elicitation instructions that request temporal sequencing, and those that do not?
2. Are the form and function of verbs produced by individuals with aphasia in discourse responses to single pictures different for:

   1. pictures that depict a complication or climax, and those that do not; and
   2. elicitation instructions that request temporal sequencing, and those that do not?

METHOD

Participants

A total of 12 adult participants were recruited through local speech-language pathologists and physicians. All had sustained a left-hemisphere cortical stroke with concomitant aphasia, and all were 6 months or more post onset of stroke (Table 1). Individuals whose profiles were consistent with aphasia of the Wernicke’s type were excluded from the study. Scores on the Western Aphasia Battery, Aphasia Quotient (WAB-AQ, Kertesz, 1982) are in Table 1, with participants ordered by WAB-AQ score. For nine participants, aphasia severity ranged from mild to moderate. For three participants, scores were above the WAB-AQ normal cutoff score of 93.8. Of these, one displayed word-finding difficulty consistent with aphasia in his spontaneous communication, and two with the highest WAB-AQ scores did not, even though they had previously been documented by their referral sources to have aphasia. Participant demographics are provided in Table 2. All participants were community-dwelling native English speakers living in the Dallas-Fort Worth (Texas) metropolitan area, and all but one were raised in the southern United States.

TABLE 1.

Western Aphasia Battery, Aphasia Quotient (WAB-AQ) scores and subscores

Participant number	Time post onset of stroke (in years)	Aphasia severity	WAB-AQ (max=100)	WAB-AQ subscores
				Spontaneous speech (max=20)	Comprehension (max=10)	Repetition (max=10)	Naming (max=10)
1	2.5	Moderate	59.5	12	8.95	2.8	6
2	4.5	Mild-moderate	77.2	18	9.7	3.2	7.7
3	12.5	Mild-moderate	80.8	18	7.8	6.2	8.4
4	5.5	Mild-moderate	81.9	17	7.95	6.8	9.2
5	4.0	Mild	85.9	17	9.05	8.9	8.0
6	3.0	Mild	87.5	17	9.25	9.2	8.3
7	1.5	Mild	87.8	18	10	7.3	8.6
8	11.0	Mild	89.2	18	10	8.0	8.6
9	2.5	Mild	92.0	18	10	9.8	8.2
10	0.5	Mild	95.2	19	10	8.9	9.7
11	1.0	(Resolved)	97.7	20	9.85	9.7	9.2
12	1.5	(Resolved)	99.7	20	9.95	10.0	10

Time post onset of stroke is rounded down to nearest half year. WAB-AQ is calculated by doubling the sum of the WAB-AQ subscores.

TABLE 2.

Demographics of participants

Participant number	Age (in years)	Ethnicity	Gender	Socio-economic status (maximum = 7)	Highest education
1	73	African-American	F	2	Community college
2	47	African-American	M	2	Community college
3	50	African-American	F	4	High school
4	41	Caucasian	M	3	High school
5	71	Caucasian	F	7	Graduate school
6	64	Caucasian	M	7	Graduate school
7	46	Caucasian	M	5	High school
8	56	African-American	M	6	Community college
9	57	African-American	F	2	High school
10	64	Caucasian	M	7	Graduate school
11	70	Caucasian	M	7	Graduate school
12	23	African-American	F	7	Graduate school

Socio-economic rating was adapted from Featherman and Stephens (1980), where higher numbers reflect higher socio-economic status.

Discourse elicitation

Stimuli.

Participants were presented with five pictorial stimuli as part of a larger standardised testing and discourse elicitation protocol. Two of these were composite pictures from standardised tests: “Picnic Picture” from the WAB and “Cookie Theft” from the Boston Diagnostic Aphasia Examination, Third Edition (BDAE-3, Goodglass, Kaplan, & Barresi, 2001). “Picnic Picture” does not depict a complication or climax, and “Cookie Theft” does (i.e., a falling boy and an overflowing sink), so the latter may represent a mid-point on a continuum between composite and complex picture types. Three additional (complex) pictures were selected for their depiction of a complication or climax. These were: (a) a Norman Rockwell print depicting a family conflict over church attendance (“Easter Morning”); (b) a Norman Rockwell print depicting a delivery truck blocked by a stubborn dog (“Dog/Truck/Alley”); and (c) a Pulitzer-Prize-winning photograph by Annie Wells of an emergency rescue of a woman from flood waters (“Flood Rescue”) (Rubin & Newton, 2001). Order of presentation of the complex pictures (“Easter Morning”, “Dog/Truck/Alley”, and “Flood Rescue”) was randomised.

Instructions.

“Picnic Picture” and “Cookie Theft” were presented according to test administration instructions of the WAB and BDAE respectively (two elicitations), neither of which explicitly requests temporal organisation in the response. For each complex picture (“Easter Morning”, “Dog/Truck/Alley”, and “Flood Rescue”), there were two separate instructions (3 pictures×2 instructions = 6 elicitations). Participants were first presented with a typical clinical elicitation (“Tell me what you see going on in this picture.”) They were then asked for a response that explicitly requested inclusion of temporal sequencing (“Make up your own story about what happened, with a beginning, a middle, and an end.”).

The presentation of the three complex pictures was preceded by a warm-up task using the same instructions, presented with a different complex single picture (the second single picture from Nicholas and Brookshire, 1993). In the warm-up, the examiner provided examples of events that might be included in a discourse with temporal progression, and participants gave responses to each elicitation. Warm-up responses of all participants included temporal progression when the instructions requested this, which indicated an understanding of the task.

Transcription

Responses were audio recorded and orthographically transcribed. Initial transcriptions were checked by a second transcriptionist, and differences in transcription were reconciled through agreement between the two transcriptionists.

Analysis

Discourse genre.

The discourse genre produced in response to each of the eight elicitations was determined, based on inclusion of features characteristic of each discourse type. Discourse was categorised as narrative when temporal organisation and agent orientation were present. Discourse was categorised as descriptive when actors and their actions were enumerated, with little or no connectivity or relationships between and among them.

Verb form and function.

All main-clause verbs, and verbs in subordinate clauses following verbs of perception or cognition (“It looks like the woman is raising her hand”), were categorised by form (tense) and function (degree of association with storyline). Form (tense) categories were: past, present, future, and non-finite. Typical narratives relate past events expressed in the past tense. Function categories (Table 3) were adapted from Longacre (1989, 1996; see also Hopper & Thompson, 1980) to represent a hierarchy of relative salience or prominence of function in narratives. Verbs/clauses higher on the salience/prominence cline are those more closely associated with “storyline” (punctiliar and/or sequential events), and those lower on the cline are less closely associated with storyline. The storyline or event line is the temporal-causal backbone of narrative structure, and is not characteristic of descriptive discourse. Existentials (e.g., “There’s a dog”) that consisted only of a nominal (e.g., “A dog”) were categorised as non-finite (form) and setting (function). These were the only instances in which utterances without verbs were coded.

TABLE 3.

Salience scheme adapted from Longacre (1989, 1996)

Function	Description	Examples
Storyline	Punctiliar, sequential happenings in text world. Action, motion, and events (cognitive or otherwise), often marked with punctiliar adverbs.	“All of a sudden they came out in a straight line. The father decided to scrunch down in his chair.”
Background	Non-punctiliar, non-sequential activities and states, often expressed in progressive aspect, often atelic, sometimes marked with durative adverbs.	“The husband and wife are having a picnic.” “Meanwhile, he was holding the branch.”
Setting	Introduces participants and props, and localises the text in time and space. Expressed by statives with “be”, descriptives, equatives, existential/locational and relational clauses. Also expressed through verbs with inanimate objects.	“There’s a few dishes on the countertop.” “The little girl was concerned.” “She was a winner.” “The woman has three children.” “Flag is flapping.”
Irrealis	Suggests possible alternative reconstruction of the text world. Potential actions/events that don’t eventuate. Expressed with negatives, modals, and future tense. Includes inceptives.	“She is going to check the sink.” “He didn’t know the family was there.” “She might have forgotten.” “He’s fixin’ to (“about to”) fall.”

Descending order of categories represents increasing removal from storyline. Other categories included in Longacre’s hierarchy (flashback; author intrusions; cohesive use of pre- and post-posed adverbial clauses) constituted 1–3% of the total codings and were not included. Their low frequency and specialised function would be best addressed with qualitative analysis in context.

Reliability.

Reliability was calculated as percentage agreement between two independent raters based on point-by-point comparison of scoring decisions. Inter-rater reliability was 96.9% for the discourse genre categorisation (descriptive vs narrative genre), based on the full sample. Inter-rater reliability was 94.6 % for the verb tense categorisation and 95.2% for the verb function categorisation, based on the productions of four randomly selected participants, i.e., 25% of the participants.

Comparisons.

Discourse genre was compared across the eight elicitations within subjects. Verb form and function were compared across the eight elicitations, first with data summed across all participants, and then within subjects. A within-subjects design is essential to control for potential inter-subject variability associated with personal discourse style, dialect (e.g., Green, 1998), and aphasia type and severity.

RESULTS

Discourse genre across elicitation tasks, within subjects

Responses by all participants were consistently in the descriptive genre on all elicitations that did not explicitly request inclusion of temporal sequencing (i.e., “Picnic Picture”, “Cookie Theft”, and the three elicitations in which participants were asked to “Tell me what you see going on in this picture”). This pattern of production of a descriptive genre held whether the picture involved a complication or climax (“Cookie Theft,” “Dog/Truck/Alley,” “Easter Morning,” “Flood Rescue”) or not (“Picnic Picture”).

Of 12 participants, 9 displayed a shift to the narrative genre on one or more of the three elicitations that requested inclusion of temporal sequencing, while 3 of 12 remained in the descriptive genre when the elicitation requested temporal sequencing (see Figure 1). Individuals who produced more narrative genre responses tended to have higher WAB-AQ scores, but not exclusively so. They also tended to have higher levels of education, but not exclusively so. Correlations could not be calculated due to small numbers and range restriction.

Figure 1.

Number of narrative genre responses produced by each participant, summed across the three elicitation conditions that explicitly requested inclusion of temporal sequence (maximum = 3).

Tense and function compared across tasks: Summed data

Group-summed data of verb tense and function compared across tasks are presented in Figure 2. On the typical clinical elicitations (“Picnic Picture”, “Cookie Theft”, and the first elicitation with each of the three complex pictures), responses displayed an overall pattern of: (1) present tense and non-finite forms with rare past tense; and (2) function composed of background information and setting-like statives. Use of non-finites was greatest on “Picnic Picture”. On elicitations that explicitly requested inclusion of temporal sequencing (the second elicitation with each of the three complex pictures) responses displayed an overall pattern of: (1) past tense with present tense, and fewer non-finite forms, and (2) functions that included storyline. Irrealis forms were the least frequent and occurred across all elicitation conditions, with the exception of “Picnic Picture” responses. Irrealis forms deal with hypothetical, possible, or uncertain states or events that have not (yet) occurred (Givón, 1984), and are thus furthest removed from the storyline function (Table 3). An example of the shift in verb tense and function across the two elicitation categories is found in the Appendix (Example A).

Figure 2.

Cumulative percent occurrence of verb codings by tense (A) and by discourse function (B) across each of eight elicitations, where T− indicates no explicit request for temporal sequence (typical clinical elicitations) and T+ indicates explicit request for temporal sequence, based on sum of coded utterances of all participants for each elicitation condition.

Tense and function compared across tasks: Individual participant information

Typical clinical elicitations.

For all but one participant, patterns of response were parallel to those seen on the group-summed data, i.e., present and non-finite forms, and function composed of background and setting. Participant 1 differed in her use of tense—she used higher proportions of past tense mixed with the present and non-finite across these elicitations, although the functions expressed were still background and setting.

Elicitations that requested inclusion of temporal sequence.

Individual patterns of verb tense and function production are described with reference to the genre (narrative or descriptive) in which they are embedded.

There were 22 instances of narrative genre production when temporal sequence was requested (Figure 1). For all but one of these responses, verb tense and function were parallel to those seen on the group-summed data, i.e., a higher proportion of past tense with storyline function, as compared to the descriptive-genre responses. The one exception to this pattern was produced by Participant 5, for whom one narrative was produced in the present tense, even though storyline function was represented in the verb (see Appendix, Example B). Future tense forms, when they occurred, were often located at the closure of a narrative, to convey the future action inferred from the action depicted in the picture.

There were 14 instances of descriptive genre productions (non-temporal organisation) when temporal sequence was requested (Figure 1). Of these, seven, produced by Participants 3, 4, 8, 9, and 11, patterned like responses to a typical clinical elicitation, namely present and/or non-finite forms, in setting and background functions. Five responses, produced by Participants 2 and 3, showed an abbreviated storyline (past tense, storyline function) embedded in the context of higher proportions of present and/or non-finite forms, in setting and background functions. An example is given in the Appendix, Example 3. Two responses, produced by Participant 1, showed a mixture of past, present, and non-finite forms, in setting and background function.

DISCUSSION

The patterns of group results and the exceptions to these patterns provide insights into the relationship between the discourse and sentence levels, and the ability of individuals with aphasia to negotiate this relationship in their discourse productions. Findings also hold methodological implications for sampling discourse production among individuals with aphasia.

The impact of discourse genre on verb production at the sentence level is seen even in this limited sample with individuals who have aphasia. The descriptive and narrative genres are associated with distinctive patterns of verb form and function. These verb form and function patterns, in turn, may reflect on the nature of coherence in each of the genres, as the verb forms and functions contribute to that coherence. In the case of descriptive discourse, one might argue that coherence would be limited, even to the point of questioning whether instances of descriptive discourse have any discourse-level organisation at all. The extensive use of non-finite forms and setting and background functions does not contribute to the construction of a coherent whole, but rather represents a concatenated listing, similar to the language one would produce in response to a series of individually pictured objects, states, and actions. In contrast, the verb forms and functions that fill a storyline function in the narrative discourse genre display a temporal-causal interrelationship, which contributes to the overall unity and coherence of the discourse. This interrelationship stands in relief against the background action and stative setting information. In future studies, a more complete account of the relative discourse-level coherence of the two genres might examine not only the verb system, but also nominal reference tracking, and the role played by conjunctions and adverbial phrases, in addition to direct qualitative ratings of overall coherence. What is clear in the meantime, however, is that the verb is conditioned at the discourse level of organisation. For this reason, an account of verb production skills of individuals with aphasia needs to consider how verbs are conditioned by genre and coherence at the discourse level. Parallel studies of verb-in-discourse-context would be especially useful in languages with more complex tense/aspect marking on the verb.

The nine participants in this sample who produced narrative genre discourse, and those who produced a descriptive genre even though they were asked to include a temporal sequence, provide some insight into the potential impact of aphasia on the ability to produce narratives. The ability to produce narratives may be associated with aphasia severity level, as indexed by the WAB-AQ: Individuals with less severe aphasia tended to produced more narratives when temporal sequencing was requested than did individuals with more severe aphasia. In addition, the descriptive-genre responses of two of the individuals with more severe aphasia displayed an abbreviated storyline in this elicitation condition, which suggests an unsuccessful attempt to produce a narrative. This finding parallels recent findings in personal narratives of individuals with aphasia, which provide evidence for an abbreviation of the narrative event line in aphasia (Ulatowska et al., 2004). Finally, the tense forms produced by the individual in this sample with the most severe aphasia included the past tense across elicitation conditions, even though past tense was associated primarily with narrative elicitation conditions for other participants with less severe aphasia. This suggests that a more severe aphasia may be associated with difficulty in management of the verb forms necessary to construct a coherent narrative discourse. Together, these findings support the notion that the ability to produce narratives may be compromised with more severe aphasia, either at the level of discourse organisation, at the level of morpho-syntactic support for discourse organisation, or both.

However, the interpretation of these data is limited by the small numbers of participants at each severity level, a confound between aphasia severity level and education level, and absence of participants who have no history of aphasia. Narrative production difficulties seen in this study may be associated with more severe aphasia, with less education, or both. The effects of individual style of response (as seen with one participant’s narrative produced in present tense) are also difficult to isolate, because of the small numbers of participants and the lack of non-brain-injured control participants. Additional research using this design with a larger, education-balanced sample and non-brain-injured control participants would be informative, as would studies that differentiate participant groups, or individual cases, by aphasia typology.

Methodologically, the current data suggest that single pictures, under traditional clinical elicitation conditions, may be limited in their ability to examine the very construct that defines discourse, i.e., the construct of coherence. Under traditional elicitation conditions, all participants in this sample produced responses that consisted primarily of listings of actors and actions, with little or no connectivity between and among them, regardless of the nature of the picture. Included in the set of traditional elicitations may be those that request “a story” based on a single picture, because the same descriptive discourse is produced with such an elicitation (Olness et al., 2002). What is traditionally elicited by single pictures may not be discourse, and thus may not be representative of the coherent discourse units required in everyday communication. In contrast, under elicitation conditions that request temporal sequencing, the current data suggest that a more coherent discourse in the form of a narrative genre may be produced. However, even these may not be typical of everyday narratives, as evidenced in inconsistent patterns of tense, e.g., some narratives in present tense, or narratives that change from past tense to finish in present or future tense. Until additional parallel studies with more naturally elicited narratives, such as personal narratives, can be conducted, clinical researchers should question the ecological validity of discourse elicitations with single pictures.

APPENDIX

EXAMPLES FOR ILLUSTRATION

Key: Participant utterances in italics. Unpronounced portions of words are placed in parentheses. INV indicates investigator utterances. Items in square brackets are interpretations by the transcriptionist or researcher, wherein “?” indicates that transcriptionist is unsure of the transcription.

EXAMPLE A

Typical pattern

A given person, in response to a given picture, produces different discourse genres, verb forms, and verb functions under two different elicitation conditions. When no temporal progression is requested (typical clinical elicitation), descriptive discourse in present tense with background and setting functions is produced. When temporal progression is requested, participant produces narrative discourse in past tense, with storyline function (underline) added to background and setting function.

Participant 4: 41-year-old Caucasian male with high-school education and mild-moderate aphasia.

Responding to “Flood Rescue”, “Tell me what you see going on in the picture.”

She’s in the water. And like she’s tryin(g) to, she’s drowning, looks kinda like drowning a little bit. And he’s got a, he’s tryin(g) to ge(t) (h)im, get her out. Well she’s on, she’s on the, on the uh limb. Tryin(g) to get (h)im, get her. But she ain’t quire [=?quite] far enough. They’re too far apart right now. (INV: Mmhm. Great.). Okay.

Responding to “Flood Rescue”, “Make up your own story about what happened, with a beginning, a middle, and an end.”

Hmm. Okay. Um she was in the water swimming. And the water got real high. And, and she started to drown. And um she, (a)n(d) then this guy come over and helped him, helped her on a, on a limb. And she finally got ahold of him, of her. And now she’s out of the water and doin(g) a lot better.

EXAMPLE B

Atypical pattern: Narrative with storyline in present tense

Participant 5: 71-year-old Caucasian female with graduate-school education and mild aphasia.

Responding to “Dog/Truck/Alley”, “Make up your own story about what happened, with a beginning, a middle, and an end.”

And this uh, the little one takes the t, tip [:trip] in the wrong direction. (INV: Mmhm.) And these three [?] people and twelve people. (INV: Yeah.) Xxx. Come up in there and, and “Get out!” And ch- coach it. (INV: Mmhm.) And he’s mad. (INV: Mmhm.) And the little girl xxx [unintelligible] sits there. (Laughs.) Uh and the middle I used that. (Laughs.) (INV: Okay. That’s the middle. Okay.) And then, uh sh(e), she, she gets him out of the way.

EXAMPLE C

Abbreviated narrative (underlined) embedded in context of present forms in setting and background function (descriptive genre)

Participant 2: 47-year-old African-American male with community-college education and mild-moderate aphasia.

Responding to “Flood Rescue”, “Make up your own story about what happened, with a beginning, a middle, and an end.”

There’s this lady in the water. And she’s young and everything. But she all a [?] sudden then got into the water. And she’s, the waves are trying to take her away and everything. But she don’t have nothin(g) around her. And uh, this uh, this fisherman no, this uh, what you call uh… [word finding attempts continue] …no he’s he’s he looks like he’s a um, fireman. (INV: Fireman. That’s good.) And he’s tryin(g) to get to her and put this brace around her body so he can get her out of the water, and res- rik- res(c)ue her.