People with intellectual disability can be supported by staff encouraging their skills in communication and in physical tasks. In a qualitative study, I used video evidence from a residential home and from 2 garden therapy services to argue that physical tasks are structurally more likely to result in successful performance (and corresponding positive assessment), whereas verbal tasks tend to result in failure (and corresponding correction and unsatisfactory interaction). I suggested 7 distinguishing characteristics of the 2 kinds of task and briefly discussed the policy implications for supporting people with intellectual disability.
In institutions that support people with intellectual disability (ID), everyday life is punctuated by episodes of interaction, many of which can be characterized as based on a physical task or a verbal exchange. Do physical tasks and verbal exchanges differ in their interactional success? Performing a light physical job and engaging in an apparently simple question-and-answer exchange may seem equally as easy, but their interactional structures—what opportunities they offer and their rhythm, scope, and timing—may provide radically different possibilities for collaboration and achievement. In this article, I address this question by means of a close qualitative study of interactions between members of support staff and people with ID as they engage in everyday physical and verbal tasks.
The question to be answered is a qualitative one—what do different kinds of task afford for interaction?—so I describe the ethnographic background of the interactions and analyze them closely by means of conversation analysis (CA). Why use CA rather than ethnographic observation on its own? The answer lies in CA's ability to identify exact interactional practices that would be difficult or impossible to capture with contemporary note taking and too subtle and fleeting to be fully recalled in retrospective interview.
The basic insight of CA is that people use the structure of conversational interaction (questions that require answers, observations that require evaluation, directives that require compliance, etc.) to achieve their interactional goals. These structures provide a set of expectations that people can use to display their understanding of how the interaction is progressing and what the previous speaker means by what he or she says. To give a brief example, CA work has uncovered the expectation that invitations are to be responded to promptly and positively. Notice what the unnamed nonresponding speaker is doing at Line 3, simply by not doing anything (Levinson, 1983, p. 320; line numbering altered):
C So I was wondering would you be in your office on Monday
(.) by any chance?
B (2.0 s)
C Probably not
As Levinson remarked, the 2-s pause is “not just anybody's pause or nobody's pause. … Rather, it is assigned by the system to [the nonresponder] as [his or her] silence” (p. 320). The normative response to a request is compliance; hence, everyone is free to deduce that C's request for an appointment on Monday has been efficiently turned down. C's own abashed diagnosis only brings the deduction into plain view.
The design of turns at talk has, over the past 40 years, developed into a multidisciplinary enterprise, attracting sociologists, linguists, and psychologists, among others (for a recent overview of CA's methods and style, see Sidnell, 2010). The result is that CA now provides a robust set of analytical tools applicable to all conversation.
Within the field of research on ID, researchers have used CA as a fruitful means of “drilling down” into the way in which staff and adults with ID communicate. Examples include the study of the communicative strategies of people with ID (e.g., Wootton, 1989), the practices of their assessment (e.g., Maynard & Marlaire, 1992), and the manner in which they manage their identities in interviews (e.g., Rapley, Kiernan, & Antaki, 1998). More recently, it has also been used to uncover the working practices of residential service staff members in dealing with people with ID during activities (Finlay, Antaki, & Walton, 2008), how people make choices (Antaki, Finlay, & Walton, 2009), and how people with ID manage interactions with personal support workers whom they employ (Williams, 2010). As yet, no close relation has been found among CA research on support practices, self-determination, sheltered activities, employment, and supported employment, but the promise of CA is that it will contribute to the understanding of how each of these key areas of interaction is successfully accomplished. The question, then, is, “Do the formats of physical and verbal tasks afford different likelihoods for interactional success and failure?”
Data and Method
I collected data in two distinct settings: garden activity centers and a residential home for adults. In the garden activity centers, staff instruct and encourage service users in a variety of simple and often repetitive gardening jobs (e.g., weeding, filling pots with compost). In the residential home for adults, staff are on hand to support the residents in daily living, and a great deal of the interaction involves staff asking residents for their accounts of what they are currently doing, have recently done, or will do in the future (e.g., describing what activities they took part in the previous day or will take part in the next).
Details of the Data Collection Sites
The data for garden physical activity came from two services, one in the south of England and the other in the Midlands (central England), that offer gardening activities to members of vulnerable groups, including adults with ID, who are referred there by local social service authorities. I call these service providers GardenSpace and Foxwood Garden Project. They provided some 20 hr of video that was shot, on and off, over a period of 2 years. The residential home, located in the south of England, I call Comber Hall Way; it provided about 30 hr of video, shot over a period of about 4 months.
I viewed the raw footage from the three sites to identify clear episodes of staff and client engaged in an either predominantly physical or predominantly verbal task. To exclude extraneous influences on their interaction, I then selected only those occasions on which the two participants were interacting only, or very largely, with each other, not significantly disrupted by distractions elsewhere. Typical physical task episodes from the garden centers included potting seedlings; pruning plants; painting; washing pots; cleaning tools; digging; stacking pots, trays, or other equipment; fetching and carrying supplies; and so forth. Typical verbal task episodes from the residential house included deciding on the week's menus; deciding on leisure activities; explaining preferences for courses of action; reporting on recent experiences; making choices around the house; explicating what the resident had just said; and so on. Although this is my assertion only, the 50 hr of video showed that the episodes I present here were entirely typical of the everyday life of the three services I visited.
The qualitative method in CA involves very carefully inspecting two aspects of the video record: the exact delivery of talk and its sequence over the interactional exchange. The principle is to identify the ways in which the participants in the scene use conversational devices to carry out their business: In the cases shown later, the focus is on how the staff members use instructions, directions, advice, and questions and how they respond to their clients' talk with assessments, corrections, and elaborations.
In all cases, formal ethical oversight was given by my own university and the centers involved, and written informed consent was obtained from staff and service users before filming began. All names and identifying details have been changed in the transcripts.
In reporting the results of the qualitative analysis, I first invite the reader to consider an example of an interaction during a physical and then a verbal task. Then I outline what I suggest are seven systematic differences in what the two formats afford for successful interaction. Success is taken to be achievement of the task (as assessed by the participants themselves), the comparative equality of contribution from each participant, and the absence of requests for repair or clarification, or repeated instructions and commands, from staff.
Consider the progress in the following interaction, in which garden therapist Carl is overseeing a small group of service users who are potting seedlings. Carl moves around the group, and at one point comes to stand by Rory, who has a volunteer beside him. (A brief glossary of the notation is provided in the Appendix.)
The real-time unfolding of the task gives Carl the occasion to do three things with Rory: to guide, enhance the meaning of, and assess his performance. The guidance, or instruction (Zinken & Ogiermann, 2011), is visible in lines marked I in the left-hand column (Line 1: “need this hand to help you,” etc.), the enhancing commentary in lines marked C [Line 12: “this one here, (I think that's a wee:d,)”], and the assessments in lines marked A [Lines 9–10: “now that's it”, (.5) “goo::d”, (1.5) “(th's) goo::d, excellent,” etc.].
The benefit of seeing the transcript is that one sees the reciprocal development between what the participants are doing. Instruction episodes such as this one, in which there are objects to be manipulated and elements to be named, afford great scope for the interplay between learner and teacher, and the details of pointing, touching, and gesturing can be very instructive (for recent work on bodily “quoting” and the use of the visible environment in instruction, see Haddington, 2010; Keevallik, 2010).
How Carl intercedes is tailored to the task of instructing, correcting, teaching, or assessing, as appropriate, exactly how Rory has done. Space constraints prevent going through the episode as a whole, but see, for example, how at Line 24 Carl intervenes in Rory's current, and presumably faulty, action to mix instruction (“just go carefully”) with an educative account (“because that's an iris”). Carl further instructs Rory to use both hands, complementing the verbal instruction with touch (Lines 29 and 32) and again giving an educative account of the meaning of what Rory is meant to do [“you (take out) the weed”].
This short episode is entirely typical of the interactions between the garden staff and the service users, especially in one-to-one activities when doing inside work (outside work, such as digging, tended to be interrupted by comings and goings of other users and staff). The user is launched on a simple, repetitive task and left to proceed; support staff come in to guide, enhance the meaning of, and assess the performance in sequence. Before analyzing in more detail the opportunities such physical task–based episodes provide, let us consider and compare what happens when the task is a verbal one.
Let us take the example of what turns out to be a reasoning task, in which the staff member requires the resident to show that he understands what to do if someone were ever to cause him harm. Support worker Kath and resident Henry are among others in the kitchen. Henry is drying and putting away crockery. Kath has turned the conversation to what Henry was doing earlier in the day. Extract 2 shows how the episode starts.
At this point (Line 20), Kath explicitly recognizes that Henry's memory of the morning's events seems to be too uncertain for further questioning, and she uses a preenquiry (Schegloff, 1988): Do you remember what we were talking to you about the other night? This establishes a different, presumably related, basis for more questions.
To summarize the episode, Kath knows that Henry attended a role-play that morning, the point of which was to educate the residents in what to do if someone hurt them. She embarks on a series of questions to get Henry to rehearse what he learned, as evidence that the lesson has taken hold. Failing to remember the morning's role-play, he is asked what the house staff has told him (presumably about the same topic) on some earlier occasion. In a series of repeated questions, Henry again fails to come up with appropriate answers, and his ultimate response is treated with disdain (“oh you'd ring the police and tell them that you'd be happy is it?” “no I don't think so”). Henry has been set a verbal task—to remember, or reason out, why he might ring the police and what it is that he would tell them—and he has failed. (Such exchanges between staff and residents were common in this service.) To dwell on just one aspect of the exchange in more detail, consider the repeated questions, marked by R in the left-hand column. Kath's commitment to building up by degrees Henry's account (of whom to turn to if he were hurt) means that she presents him with a series of questions that follow a trajectory (that projects a sequence, in CA terms; Schegloff, 2007) that she understands, but that he might not. In each case, the repeated question signals that Henry's answer, or more usually his silence, was unacceptable (was a turn that was not formatted to terminate the sequence) and that the interaction could only move forward once the question had been dealt with. Recall the example I gave in the introduction: An invitation projects a satisfactory response, and if one is not forthcoming, then the inviter draws an implication. Here, Kath's question projects a satisfactory response, and Henry gives (what he intends to be) one; that in turn projects an acknowledgment and a move to next question or next topic. Kath's turn is very markedly not a move onward (Kath reissues the question), so Henry must infer that his answer was wrong.
There is a strong flavor here of an extended series of test questions (Edwards & Mercer, 1987) to which the questioner already knows the answer. Such an asymmetry of knowledge is standard in the schoolroom or at home between adult and child. The questioner sets the pace and arbitrates the correctness of the response. The purpose is educational, but the maintenance of the sequence of question and answer (see Schegloff, 2007, on how sequences can be extended over multiple turns) is in the hands of the questioner. The answerer's task is complex, insistent, and difficult to see the outline of.
Identifying Seven Interactional Benefits of the Physical over the Verbal Task
The two vignettes allow me to suggest two sets of features that are afforded by the different formats. The physical activity format provides for easy, repetitive tasks, with visible requirements, for which the service user can at least make an approximately successful start; thereafter, her or his actions can be encouraged, fine-tuned, enhanced, and appreciated. However, the format of verbal tasks (explaining things, giving judgments, recounting events) is, or can be treated as being, complex actions that will be well or ill formed according to the questioner's criteria, with little prospect of being guided during them; such tasks will be more difficult from the start, because they cannot easily be broken down into subcomponents without risking confusion. Putting the characteristics of the two kinds of task in tabular format, and at the risk of exaggerating the differences, one sees the benefits as laid out in Table 1.
Look at examples from the physical activity data (with only glancing comparison with verbal tasks, for reasons of space) that support each of these claims. Recall the qualitative claims that I am making: I am not counting how often physical tasks and verbal tasks are presented to the clients (although that would be an interesting further project, if I succeed in establishing the differences between them) but rather what the two kinds of task provide for in terms of productive interaction. It is important to bear in mind that these observations are limited by the episodes available and by the specific combinations of staff experience, institutional imperatives, and local cultures; I cannot claim they are universal without more extensive study. The headings I use are best approached as a heuristic list than as a definitive catalogue.
The kind of tasks I saw in the data were easy and could be made still easier at the outset in view of the likely problems that the service user might be expected to encounter. For service users who find it especially difficult to perform motor tasks, this advantage was of special benefit because the range of what they could do was necessarily limited (and would be still more so, of course, if the task were a verbal one). For example, volunteer Danny is about to set service user Clark on the job of varnishing some pots that had been painted. Danny knows that Clark has comprehension difficulties and is awkward with his hands; hence, Danny introduces Clark to the task by carefully holding the paintbrush in such a way that it affords him the obvious and easy requirement to grasp it, ready for use.
Physical tasks can be left in the hands of the service user, with the staff member on hand to intervene only when necessary; such unilateral activity within an interaction is, of course, not possible if the task is a conversational one (the speaker would seem to be ignored or tuned out by the others). For example, consider a staff member and two service users, each concentrating on their own pruning job. They share an intersubjective space—each is available to the other for communication—but each can pace his or her progress on the task independently.
A physical task, especially a repetitive one, may be suspended to allow time out for a variety of reasons—fatigue, loss of concentration, distraction—without necessarily imperiling the successful performance of the task as a whole. In another example, service user David turns to the videographer, secateurs in hand, ready to return to the pruning task at hand, and engages her in a brief exchange about what he will be having for supper in the evening. David can suspend his task without sanction or risk of compromising the ultimate achievement of his objective (pruning the leaves on the plant in the pot). In a conversational exchange, there is no such time out—every absence is an attributable and accountable one (Sacks, Schegloff, & Jefferson, 1974).
In the typical physical task in these episodes, the service user was engaged in a series of actions that had a sequence, building to the completion of a cycle and repeating it, which means that the watching staff member had a constant index of where the service user was on the trajectory of the task and could identify things going awry at an early stage. This is not normally possible when a service user is performing some verbal task, for example, giving an account of the previous day's activities or expressing a preference for a holiday destination. In such cases, the beginning, middle, and end of the task's performance are not clearly predictable or visible, so the staff member may be less sure of whether things are going wrong until it becomes so obvious that intervention takes on an explicitly corrective and, by implication, fault-finding sense. Certainly, if one turns back to Extracts 2 and 3, Henry's awkward attempts to explain why he should call the police have too little internal structure to allow for easy monitoring and early guidance; perhaps it is for this reason that Kath's questions have such an interrogative edge and her treatment (“oh you'd ring the police and tell them that you'd be happy is it?” “no I don't think so”) of his ultimate answer appears to be so scornful. Compare that with the following example of a physical task.
Garden therapist Sophie was supervising service user Frank in arranging some potted geraniums to water them. She modeled where to put them, and the watering can was within reach, at their feet. Frank stood and looked at Sophie.
Sophie set the watering task in action at an earlier point. The first phase consisted of Frank's arranging the geranium pots at a certain place. Once this was done, he would water them; the watering can had been filled and fetched and was ready at their feet. Frank duly set out the geraniums and then straightened up, but instead of proceeding, he looked toward Sophie, whose attention was momentarily elsewhere. She then turned and, in view of Frank, looked down at the now set-out plants; then she gazed directly at Frank for about 1 s; after which he stooped down to pick up the watering can and began watering. Her monitoring of where he is in the task, and her silent prompt, provide a no-fault guide to get Frank back on to the task at hand.
Opportunities for Coaching
Verbal exchanges provide staff with opportunities to take what the client has said and use it as a basis for some educational work, but at the cost of implying that the client's utterance is interactionally deficient in some way. A physical task, however, allows the staff member to seemingly build on the client's performance and superimpose a layer of extra meaning without implying that the original layer was inappropriate. In the following exchange, staff member Mickie is supervising a group of service users in filling pots with compost.
Mickie's invitation (“who's going to show me the best way to fill a pot”) is an invitation that nominates any member of the group as the next speaker. Such nomination is unusual in noninstitutional talk but has the advantage in teaching settings of attributing all those present with the right to respond: that is, not only to perform the task but also to demonstrate it. David volunteers, but notice that his initial efforts at preparing the pot involve only (rather laborious) shaking out of debris. Mickie's gloss (“thassit, you've checked the pot to make sure there's no cracks, yeah?”) takes the opportunity to recast what he has done as checking for cracks, which upgrades his performance and provides a form of embedded coaching (on a par with the embedded repair of conversational errors noted by Jefferson, 1974). At Line 17, she introduces a little tip that delivers what she had nominally asked the group to provide (“the best way to fill a pot”). Once she has secured David's supposed confirmation of this advice, she moves (Line 33) to cast the entire episode as one that speaks to David's, rather than her own, competence: “that was very good (.) d-David, good demonstration.” Seen out of context, on its own, this would be a simple piece of praise; in context, one sees that the praise is a vehicle for a reformulation of the staff member's own work into a measure of the service user's competence.
Unlike the requirements of an account or explanation, or some piece of reasoning and so on, the end product of a physical task can be unambiguously clear. For example, a staff member asked service user David to follow her example in moving some flowerpots out of the flowerbed and onto the path. She did the first few while he watched, then stepped back and said, “C'mon you do those last few for me.” He could quite clearly see where they had to end up, and he completed the job with no fuss.
Unlike such visible endpoints, which can be pointed to, are available at any time, and require (at worst) only simple imitation, giving someone a verbal task necessarily means that the endpoint is abstract, invisible, and not available for imitation. Indeed, when a user repeats some or all of a staff member's question in trying to provide an answer, this repetition risks being interpreted as mere echoing, and the staff member may repeat or reformulate the question (for an account of such troubles, see Finlay & Antaki, 2012).
A physical task is manifestly something that has to be learned and would have to be learned by anyone. Talking, however, is normally thought to require no special skill. So one can be held less accountable for failure on a physical task than for failure on a verbal task, and it occasions less repair work when it happens—indeed, in much of the data, service users' obvious inefficiency, and occasional downright ineffectiveness, in performing even simple tasks passed unremarked by support workers. For example, wheelchair user Nicholas was asked to dig up a deep plant using a garden fork. In fact, he could not use a force adequate to drive the fork sufficiently into the ground to get under the plant, nor would his position in the wheelchair allow him to lever it out if he could. The volunteer came around to Nicholas's right and took over the task, with no averse comment.
Failure at verbal tasks (which were very common in the Comber Hall Way residential data), however, usually calls forth some kind of palliation or repair on the one hand or, conversely, a seemingly aggressive pursuit of a proper answer to the staff member's questions on the other (again, see Kath and Henry's exchange in Extract 3 for an illustration; but see also Antaki, Finlay, Walton, & Pate, 2008) for more examples and analysis of the phenomenon).
In this article, I set out to chart the different affordances that physical and verbal tasks offer people with ID in interaction with those who support them. I used video data to show that a repetitive physical task, such as putting seedlings into pots or pulling weeds, has more structural opportunities for positive interaction than real-time verbal tasks (explaining, giving an account, etc.). Physical tasks offer what Vygotsky called scaffolding, which staff can use to build up the client's confidence and competence. The value of CA is that it shows in some detail how this scaffolding is done, and it might form the basis for staff training (for a recent exploration of Vygotskyan ideas in CA, especially how they actually play out in interactional practice, see Gardner & Forrester, 2010).
Some caveats are in order. I can say nothing about the inner psychological benefits of work or talk, at the time or over the long haul—the feelings of satisfaction and empowerment in seeing a job well done and so on. My analysis is strictly limited to the visible occasions provided for independent work, constructive dialogue, and successful performance. My warrant for the benefits of work or talk is not based on psychological measurements but on how the interactions come off. The validity of the claim that an interaction ends up badly or well lies in the public conduct of the participants involved. Recall the interaction between Kath and Henry, who was asked to explain the procedure for reporting abuse. The interaction was dominated by Kath's questions, which, by pursuing the same points over many turns, implied that Henry's answers were inadequate and unacceptable, and the episode ended with Kath mocking Henry's ultimate reply as he departed the scene. In the examples of physical activity from Rory's planting (Extract 1) onward, one sees a less unilateral domination of proceedings, greater achievement, more praise, more settled work, and interaction unruffled by the staff member's insistent demands for an adequate reply. Such observations all want further specifying, of course, but they leap to one's attention, even in raw form.
Physical and verbal tasks are not exclusive. It is important to stress that the staff in the garden activity centers did not merely launch clients on tasks, monitor them, and positively assess them, with no intervening conversational exchanges. There was plenty of talk. However, when the talk was not geared to what the service user was doing at the time (there was much prior explanation and commentary before and after tasks, and casual chat at any time), it was difficult for the staff to gauge just how much was being taken in. The garden tasks—whatever their horticultural benefit—were a means by which the service user's understanding and commitment could be made visible. They gave the staff member something concrete to build on.
I am certainly not saying that verbal tasks—explaining, accounting for oneself, reasoning out a problem—provide no interactional benefits. I do think nevertheless that the data show that the structural affordances of such tasks are simply less promising than those of physical tasks. That does not mean, though, that their limitations cannot be overcome, at least to some degree, by very careful calibration of just how difficult it is for the client to make the opening move, how to intercede without implying fault, and how to enhance the client's understanding of what he or she is doing without confusing the situation. All these can be done, but they are simply harder to carry out, given the real-time, normative demands of conversational exchange. An interaction based on a physical task has significant interactional opportunities that will deliver at least some of the things that conversation would find more difficult: guided performance, educative enhancement, and greater scope for positive assessment.
I shot the GardenSpace and some of the Foxwood Garden Project footage. I'm very grateful to Emma Richardson, who shot most of the Foxwood Garden Project footage and to Chris Walton, who shot the Comber Hall Way footage on a project supported by Economic and Social Research Council Grant RES-148-25-0002.
Editor-in-Charge: Stephanie Dean
Charles Antaki (e-mail: c.antaki@Lboro.ac.uk), Department of Social Sciences, Loughborough University, Epinal Way, Loughborough, Leicestershire LE11 3TU, United Kingdom.