The interactional organization of aphasia naming testing

Abstract

In this article, Conversation Analysis (CA) is used to investigate the nature of aphasia naming tests in terms of their properties as a specialized form of social interaction. The basic test-item sequence which occurs in these tests is shown to be made up of a three-part sequential structure consisting of (1) a testing prompt, (2) a proffered answer by the testee, and (3) an acceptance or declining of that proffered answer by the tester. A declining prompts a further answer to be proffered, and this cycle continues until either an answer is accepted by the tester or until the participants treat the testee as being unable to produce the relevant picture name. It is suggested that the results of the analysis have implications for understanding naming tests as instruments which generate theoretical and clinical findings through particular talk-in-interaction practices.

• Aphasia
• conversation analysis
• naming testing

Notes

*For a list of the main CA transcript symbols, see the Appendix. A fuller list is available in Schegloff (2007). It may be worth noting here the inverted question mark symbol “¿” which marks a rise in intonation which is greater than that marked by a comma, but less than that marked by a (non-inverted) question mark. While this symbol tends not to be widely used within CA publications, it is used in several of the transcripts in this article (including in lines 12 and 15 in Extract 4) to capture the intonation often used by these testees in producing an answer.

†For the notion of “positions” in a sequence and the distinction between “turns” and “positions” see Schegloff (1992). Here, the relevance of the notion lies in the fact that answers by PWA, or receipts of those answers by testers, may occur later than the second or third turns, respectively, following the testing prompt but are still produced to be heard as systematically sequentially related to it (i.e. in second or third position, respectively). This “delayed” production of the turn can be due, for example, to “intervening” talk or actions being produced prior to the answer or the answer’s receipt. An example can be seen in Extract 6 below. There, the answer is produced as a second position response to the picture prompt, despite that the fact the picture prompt was first oriented to in line 2 (when the PWA turned over the page and looked at the test-item picture) and that several “intervening” turns by both the PWA and the tester have been subsequently produced prior to the answer in line 7.

*See Marlaire and Maynard (1990) and Maynard and Marlaire (1992) who discuss a collapsed form of the sequences in their psychological test data, though in that case it is the turn in third position (e.g. an acknowledgement) which is not produced. Maynard and Marlaire (1992) suggest that the completion of the sequence is marked instead by the clinician pausing and producing a new prompt. The non-production of a third position response appears to mark a systematic orientation by the tester to a certain interactional contingency; as Marlaire and Maynard (1990, p. 96) note: “The third slot is dispensed with when the child’s attention and answers are proper and correct; it is reinvoked when they are not”. As will be seen in this article, testers in these aphasic testing data can also cease to produce third position turns. This can co-occur with the non-production of first position turns, leading to a “doubly collapsed” sequence (see Extract 12).

†Despite the fact that (after the first item or two of the test) it is regularly the testee who initiates an orientation towards the next picture, the interactional relevance of that picture as an object to be named can be seen to have been set up by the tester from when he or she introduced the testing activity and then provided a verbal and/or gestural orientation to each of the first few pictures. In Extract 4, for instance, the activity is introduced as a “series of pictures” to be named (line 1), setting up the relevance of naming not just the first picture but rather all of them in the test.

‡The target is “bear” but the tester marks “polar bear” as correct on the scoresheet.

§For ease of transcript readability, in subsequent transcripts non-vocal features of the interaction, such as pointing and eye gaze, will not typically be presented, although they will be included when pertinent to the phenomena under discussion.

¶The terms “evaluation” and “assessment” are both used to refer to this type of third-turn response within testing sequences (Schegloff, 2007). I will henceforth use “assessment” as the default term.

*It should be noted that a tester accepting an answer as an action within the interaction is separate from that answer then being scored as correct. There are examples in the data of a tester accepting an answer and scoring it as incorrect. Why a tester will choose on some occasion to accept an incorrect answer rather than decline it and prompt a further attempt by the testee cannot be further explored here.

*Relevant aspects of conversation-analytic work on preference organization will be discussed below. For an overview of preference organization see Schegloff (2007).

†The error by the PWA here may be a visual one, confusing the picture of a mountain for that of a glacier. This possibility may influence the lexical form of the mitigated declining produced by the tester here (“could be”).

*This display of relative certainty is well-placed since all eight answers here are indeed correct.

†Similarly, acknowledgment tokens such as “mm hm” in the post-answer slot can be equivocal, since they can function both as acceptances (Extract 4, line 9) and as precursors to declinings (Extract 10, line 2–6; Extract 11, lines 5–8). Again, how an acknowledgement token might be hearable as one or the other on any particular occasion will demand further analysis.