Abstract
This paper looks at the use of brainstorming as a method in two successive engineering meetings. Through a close analysis of engineers' interactions I examine how brainstorming rules are used, and often ignored, in the course of the sessions. It is found that engineers rarely explicitly orient to the rules of brainstorming, and that many sequences of interaction appear at first glance to be in breach of brainstorming rules (but are not censured within the meeting as such). Through a set of empirical examples, I develop the case that engineers are principally orienting to social order – the ‘rules’ of social interaction – in preference to the rules of brainstorming. This finding has a number of implications for design research: (a) it enables a reassessment of the nature and use of methods in design, (b) it reveals what other ‘rules’ are in play and found valuable in design activity, and (c) it uncovers aspects of the organisation of collaborative idea generation that have not been previously identified.
1. Introduction
Design methods have been a central topic in the historical development of design research. For some researchers, methods for doing design offered a way of formalising design practice. This was seen as a key to understanding how designers work (a way of describing designers’ activity), to automating design (e.g. by means of algorithms), and to educating design practitioners. In many ways, Simon (Citation1981) remains an archetype of such research programmes and their ambitions. Over time, however, methods were challenged as a means to each of these ends. Problems were encountered when attempts were made to account for the activities of designers in terms of the steps prescribed by design methods, e.g. Hales and Wallace (Citation1988), the inherent limitations (philosophical and practical) of formal logic and its derivatives to produce solutions to genuine design problems has been the basis of a wide-ranging discussion in design research (for example, see Rittel and Webber Citation1973; Simon Citation1973; Schön Citation1990; Winograd Citation1991; Coyne and Snodgrass Citation1993), and the very idea that good design work is, or can be, the straightforward outcome of the application of a method was not something ever vindicated by the results of methods-based design programmes at universities and design institutes. Other bases were sought for design education (e.g. Schön Citation1987). Whatever it does take to be a successful designer, it is surely more than having facility with design methods. Nevertheless, methods are still taught, still found useful, and remain a fundamental component of many educational programmes and professional design studios.
In this paper I attempt to engage this debate from a slightly different angle, one that begins with a consideration of designers' situated and practical employment of a method. (In this respect, I am preceded by, and indebted to Bucciarelli's (Citation1994, pp. 151–164) insightful analysis of an engineering team's use of the Pugh method.) Brainstorming is a widely used method for generating new ideas that hardly needs introduction. Although there are local variations in how brainstorming is conducted, it is founded on a number of basic ideals, including suspending critical judgement during the session, allowing one speaker the floor at a time, encouraging ‘wild’ ideas, and building on the ideas of others. It is in common use in countless design consultancies, including some of the most prestigious (such as IDEO). Existing studies of brainstorming have predominantly focused on its effectiveness as a method of generating ideas (e.g. Jablin Citation1981; Dennis and Valacich Citation1994; Sutton and Hargadon Citation1996 evaluate it in terms of other instrumental payoffs of its use). My interest here is not in evaluating the method at all, but simply to inspect its use in order to see what considerations are in force when designers use and/or ignore the tenets of the method.
The data I am examining here consist of video recordings of two brainstorming meetings conducted three days apart at a product design/engineering consultancy in the UK. Both meetings were about 100 min in duration and were among the very first work meetings on the project. The task they are working on concerns the design of a digital pen apparatus that is able to elicit different heat-sensitive colours from a unique writing surface. The first meeting was about the physical and mechanical aspects of the design; the second meeting was primarily about electronic and software elements of the product. The technical details of the project are not of particular importance for the analysis I will present as my focus is on the organisation of the session more than the content of the discussion.
2 Analytical approach
The analysis I present here is informed by ethnomethodological conversation analysis, an approach that has been adopted with different focii in previous studies of designers' work (e.g. Luff and Heath Citation1993; Bowers and Pycock Citation1994; Button and Sharrock Citation2000; Matthews Citation2007; also Oak (this issue)). Since the seminal work of Harvey Sacks in the 1960s and 1970s (much of which was published posthumously), conversation analytic studies have shown how speakers and listeners co-produce a normative social order in interaction. An obvious example of social order in conversation is the expectation created by the asking of a question. That is, in a normal conversation, the next turn at talk following a question will be scrutinised (by the conversationalists themselves, e.g. the question-asker) for its answer-ness to the question just asked. The relevance of ‘next turns’ to immediately prior ones is a normative matter that conversationalists themselves exhibit in how they orient to each other in talk. That this is so can be seen by consideration of many facets of conversation, including just how delicate topic shifts within conversations actually are for conversationalists (see e.g. Sacks Citation1995, Lecture 5 Spring 1970 and Lecture 9 Spring 1971). This is just one example of the emergence of social order in interaction, but it is one that has application to the analysis that follows. One important aspect of the programme of conversation analysis has been its demonstration of how finely detailed and richly textured is the production of social order. For example, very small silences in conversation (tenths of seconds) can be shown to have interactional relevance – they may, for example, prefigure misunderstandings or other ‘trouble’, or they may mark a place for speaker transition. The point is that social order is a very finely organised phenomenon, and is observable in the details of interaction.
There are a number of ways in which such observations might prove relevant for design research. In the context of the data of these engineering brainstorming meetings (E1 and E2), there is the potential to identify possible intersections between different ‘orders’. The purpose of this paper is to examine the relationship between the rules of brainstorming and social order. I will do this by juxtaposing of some of the ‘rules’ for conversational interaction with the rules for brainstorming to see how they play off each other, as in how they intersect for the participants in the meeting.
At first glance, there are several possible intersections of social order and brainstorming rules. Firstly, the normativity of relevance in turns at talk appears that it may conflict to some extent with the brainstorming injunction for novelty. While there is an expectation in conversation that talk should be relevant to the immediately preceding talk, there is also the brainstorming injunction that ideas should ‘break existing mindsets’ as Alan instructs the group in E1 (E1, line 31 of the transcript), that divergence not convergence is desirable (E1, lines 40–41), and that all possible avenues of ideas should be explored (E1, lines 41–42). Secondly, we might expect some overlap between how interruptions arise and are handled in interaction and the brainstorming rule not to interrupt the current speaker (E1, 15–31). Finally, it may be instructive to compare the organisation of criticism in interaction with the brainstorming rule not to criticise ideas in the session including one's own ideas (E1, 7–8). These are not hypotheses so much as circumstances in which an analyst might be able to see how engineering designers manage their accountability to both social order and the rules of brainstorming. It is these three ‘intersections’ that are the focus of the study presented below.
One principal of this kind of analysis is that analysts refrain from imposing their own order, categories, structures, interests, etc., onto the data; instead they attempt to analytically recover the participants' understandings and orientations as they were displayed in interaction. One means of doing this is by inspecting the data on the basis that each turn at talking displays the speaker's understanding of what is going on at that moment. Each utterance in the transcript can be seen as an analysis of the preceding stream of events – as the speaker's own in situ analysis – and its study can direct us towards a recovery of the understanding that the participants' possessed and exhibited in the course of the interaction. A short example ( ) from the transcript of the first meeting is illustrative.
Extract 1. Speakers' turn-by-turn understandings.
Tommy's utterance at 1416 displays an analysis of Alan's prior turn. Tommy frames this as an opinion (‘I think that …’) that questions the relevance of Alan's comment to this meeting by suggesting that the discussion belongs in Monday's upcoming meeting. In a similar way, Alan's turn at 1417 acknowledges and appears to accept Tommy's suggestion, also displaying Alan's understanding of Tommy's turn as a suggestion to drop the topic.
3 You're not supposed to be negative!
At the beginning of the first of these meetings, Alan (the project manager) introduces the meeting as a brainstorming meeting, and outlines the rules of the session. In the first few minutes of the session he instructs the team that their aim is for divergence, not convergence, that they are looking for ideas that ‘break existing mindsets’, and that all possible avenues should be explored. He reminds them that they should not criticise ideas (even their own ideas), and he twice instructs the team that they shouldn't interrupt each other.
In this meeting, the first rule introduced is the rule not to criticise others' ideas, and Alan declares that it is the most important rule of brainstorming. However, apart from its appearance here at the start of the first meeting where it was listed along with other rules of the session, it is very seldom that the participants explicitly invoke any of the rules of brainstorming during the meetings. The following sequence ( ) from E1 is a rare case, occurring a little over an hour into the first meeting.
Extract 2. Invoking the rule not to criticise ideas.
At E1 1359, Alan produces a version of the ‘no criticism’ rule, following Rodney’s turn in 1356. Again, reference (oblique or otherwise) to one of the rules of brainstorming during the session after Alan's introduction where the rules were established was very rare. Looking at this sequence, we may be able to see what some of the consequences of invoking a rule in interaction actually are. Alan's use of the rule at this point paints Rodney's earlier turn as a violation of it; not only that, however, it also censures Rodney for having transgressed the rule (c.f. Wieder Citation1974). As an analyst, it is difficult to see exactly what triggers this particular censure from Alan, as there are numerous other comments made at other times that appear equally ‘negative’, but which patently did not occasion a reiteration of the rule not to be critical. It may be noteworthy that following this censure from Alan, the transcript does not record Rodney attempting another turn at talk for more than 500 lines of transcript, or about 25 min of the meeting. This may also lend significance to the scarcity of occasions that brainstorming rules such as ‘don’t be critical' are invoked – that participants themselves are aware of the possibility that censuring criticism closes down dialogue much more effectively than permitting the occasional critical comment. This raises the possibility that not invoking the rule in the meeting is actually in keeping with (the spirit of) the rule. That is, if designers appreciate that the point of the rule ‘don't criticise’ is for dialogue to remain open so that more ideas can be aired and more participants can contribute, and if the censure of a participant risks ostracising him or her, then disattending to the rule may actually be a delicate means of keeping it. And such an analysis suggests that if these designers are adhering to the rules of brainstorming, they are not doing so in obvious ways, but in ways mediated by other subtle concerns. Much more can be seen to be going on in the details of interaction than designers simply following the rules of brainstorming or transgressing them, and the details of interaction can be inspected to see what some of the participants' other concerns might be.
4 The grammar of an incomplete thought
In ordinary conversation there are some remarkable structural regularities that have been the focus of conversation analytic research over the past 40 years. One of these has to do with the places in speech that overlapping talk occurs. There are particular and remarkably consistent places in the current stream of talk where a listener will try to begin a turn if they want to speak next. In spite of the fact that conversation is a distributed affair involving multiple parties, each of whom speak at one time or another, there is surprisingly little overlapping talk. Speakers shift from one to another, but in a highly regulated way that exhibits (and allows for) remarkably little simultaneous talk. Conversation analysts have identified a number of these turn-taking (or speaker-transition) rules (Sacks et al. Citation1974), but the important thing to note here is that listeners often wait for the speaker to end a thought, or to possibly end a thought, before beginning their own turn. For instance, pauses at the end of a sentence often invite speaker transition, whereas pauses at other places in speech, such as after an infinitive marker but before the verb, as in ‘I think it's best to … ((pause))’ typically do not. Were a listener to begin speaking at one of these points of ‘maximum grammatical control’ (Schegloff Citation1996), their turn would be susceptible to be treated as an interruption by the other parties in the conversation.
That ordinary conversation has its own rules for overlaps and interruptions, and that brainstorming has the rule to avoid interruptions becomes more interesting when we see how these ‘rules’ intersected in these engineering meetings. For instance, in the following excerpt 3, Patrick appears to get interrupted twice.
Extract 3. E2, Patrick is interrupted.
At line 1064, Patrick's thought is unfinished when Sandra begins speaking. At the end of her question he tries again, starting his turn by repeating the same phrase ‘could you’ that got cut off the first time, but this is overlapped by an answer to Sandra's question, and he again does not get to complete the thought. After a rapid exchange, Jack introduces a question that appears to be his own continuation of Patrick's now twice-aborted attempt, as it starts with the same three words as Patrick's turn in 1066: ‘could you have …’. It is on account of Patrick's repeated attempt to ask the same question that we can infer he treats Sandra's intervening turn as an interruption, i.e. that he didn't get to finish his initial turn, so he tried again. This is one way an analyst might identify interruptions in the transcript. Another is simply to see where thoughts are left unfinished and how they are treated by participants in the interaction.
It was in the course of looking for interruptions that I came across an interesting feature of the organisation of turn-taking in these meetings. That is, there were a surprising number of unfinished turns – turns that did not complete a thought: a sentence, clause or phrase the way that Sacks et al. (Citation1974) found. A number of ideas just seemed to stop short, to be left unfinished. Sometimes they were completed by another participant, sometimes they were interrupted by another speaker and not returned to, and sometimes they were simply abandoned by both the speaker and the group. Initially, I thought that this signalled that there were many interruptions in the meetings (and consequently, multiple breaches of the brainstorming rule not to interrupt). As such, I expected that these unfinished turns might be treated as interruptions by the participants when they happened. But that expectation was not realised when I looked at the data. The interesting feature was that even when speakers shifted in the middle of a thought, these were manifestly not treated as problematic or interruptive speaker transitions by the participants. Two examples follow in and .
Extract 4. E1, An unproblematic speaker transition in mid-thought.
Alan's turn is unfinished in line 297 when Tommy comes in with a question and candidate answer that is also left unfinished. Todd adds a counter suggestion with a tag ‘isn't it?’ and the discussion continues. Neither of these two interjections, e.g. speaker transitions before a finished thought, is treated as an interruption by the participants. For example, they do not give any indication that they hear Tommy's or Todd's turns as an interruption – there are no repeats of sentences that were cut off, nor does the ‘interrupted’ speaker attempt to regain the floor at the next possible end-of-thought.
Extract 5. E2, Another unproblematic transition in mid-thought.
In , Tommy's turn at 1023 is of interest, which is cut off at the beginning of a phrase starting with a preposition ‘which …’. Sandra creates one continuation of the thought by talking about a library of patterns that could be scanned in, but this thought too is unfinished in the turn. Patrick takes up an extension of this notion, interrupting Sandra with a token of agreement ‘yeah’, and developing the idea into a book. It is noteworthy that although we have speaker transition taking place at a juncture in the turn other than a paradigmatic transition relevant place, the ‘interruptive’ turns at 1024 and 1025 are again not treated as interruptions by the participants.
Taken together, these examples suggest that turn transition in brainstorming meetings may not exactly replicate turn transition in ordinary conversation. Footnote1 More interesting for design research is the indication that leaving turns unfinished is a practice that co-opts others into idea generation. These examples show that this is one recurring consequence of leaving ideas hanging in the middle of a turn. Stopping or trailing off mid-turn like this can induce others to add to or alter the idea in potentially significant ways without the speaker needing to make any explicit request. No question is asked, no overt invitation to comment is made; rather, this is simply achieved with the grammar of an incomplete thought.
5 Relevant topic or new idea?
One enduring feature of ordinary conversation is that there is an overwhelming normative expectation to speak relevantly. In conversation, each turn is scrutinised by conversationalists for its relevance to the last. This is exhibited through and through by the work that speakers do within conversation to close down current topics and introduce new ones. It is also plainly evident that conversationalists request clarification when the relevance of a speaker's turn is not apparent. Furthermore, it is observable in the work (such as prefacing a story) that speakers do to introduce a coming topic as relevant when they anticipate that its relevance may (otherwise) be questionable.
McHoul and Rapley (Citation2003) use an unpublished draft manuscript of Harvey Sacks' to illustrate a pragmatic use of the phrase ‘I just had a thought’. In their analysis of this segment of Sacks', the phrase was used as a way of linking the story to come to the immediately previous stretch of talk. It served to mark the coming account as being ‘on topic’; it is a way of suggesting that what the speaker is about to say, which may first appear to be irrelevant to the current topic, actually has relevance (McHoul and Rapley Citation2003). The work that speakers do to demonstrate relevance is significant – speakers are accountable to fellow conversationalists for the particular appropriateness of their contributions (see Sacks Citation1995). In conversation there is an expectation that each ‘next turn’ should be relevant to the last with its relevance displayed in the turn itself or, if not, that its irrelevance will be accounted for (also) within the turn itself. The use of ‘on topic’ markers is one way that speakers accomplish the relevance of their talk.
In the transcript we can see a use of a similar construction employed as a topic marker, though a little more elaborately. About 20 min into the first meeting, Todd talks about a ghost toy his son has; shows how he gets to start talking about it. Apparently, in preparation for this meeting participants were asked to bring objects/products that might be of use in the brainstorming session. At this point in the meeting, Alan asks the group if anyone has brought anything else.
Extract 6. E1, Todd's use of an ‘on topic’ marker.
Todd responds to Alan's request (at 391) with a joke about only having brought a magic marker. He then uses the marker as a means of discussing an idea related to an object he has not brought. The connection is made by use of ‘it made me think uh- of a toy’ (lines 396–397). Todd begins his idea with an ‘on topic’ marker – suggesting that this idea relates, at least in ‘toy-ness’, to an earlier turn (the magic marker). The phrase ‘it made me think’ is itself interesting. It is a passive-voice construction which suggests (at least metaphorically) a causal link between the preceding object and the idea Todd is introducing – ‘it made me think …’. This phrase defers the responsibility for what Todd thought on account of what the magic marker has made him think about (this is an observation I owe to Max Eckardt). And it is in this way that he creates a link for the rest of the group between his idea and his preceding turn. To claim that one thing has made one think about something else is to claim an association between the thing and what was thought. This is one way in which we can see that participants orient to the ‘topic relevance rule’ of conversation – that they actually do this kind of work to introduce new ideas with reference to the local context of talk. This appears to be the case even to the extent that Todd here creates a local context himself in order to get to talk about what he wants to. By producing an ‘uninteresting’ object (magic marker), Todd creates a way of having an answer to Alan's request (concerning what objects people brought to the meeting); this conversational move gains him the floor. He then uses the marker to provide a link to the cereal box toy that he proceeds to talk about.
Shortly following this is another example, shown in . Todd has been sketching and describing this particular toy. Note Sandra's remark at E1 419.
Extract 7. E1, Sandra's use of an ‘on topic’ marker.
Consider ‘that's interesting cos I was thinking pastry brush’ (line 419). We might suppose that a different beginning (such as ‘what about a pastry brush’ or ‘did you think about a pastry brush’) would not do the same work of showing the relevance of Sandra's idea to the preceding talk. But ‘that's interesting’ offers a positive assessment of the previous idea, and the ‘because I was thinking’ projects a link to what is about to appear. Continuity with the preceding topic is being handled here.
Of course, one may ask how a pastry brush is actually relevant to Todd's son's ghost toy. Interestingly, this particular relevance is not demonstrated by Sandra – instead, she switches from a pastry brush (something which is relevant to one aspect of the task, namely a product that follows an uneven contour, but not clearly relevant to the toy Todd has been talking about), to a pastry cutter, then to a line-marking machine. These latter two share an obvious principle of operation with Todd's son's toy, and the ultimate relevance of Sandra's turn is not questioned by her co-workers. What is also of note is that the idea of the pastry brush, which is actually a novel idea that these engineers have not yet considered as a way of addressing this problem in the meeting, is not pursued by Sandra or the others. Rather, the conversation continues along a more obvious extension of the current topic, concerning the use of wheels to guide the print head.
One further example, , shows (a) that designers also explicitly orient to the topic-relevancy of their contributions, and (b) the amount of work designers actually invest in order to successfully introduce a discontinuous topic or idea.
Extract 8. E1, The work required to introduce something ‘off topic’.
Todd, after a number of false starts, draws attention to the ‘off topic’ (line 949) character of the coming idea. The multiple pauses, cut-offs, uhms, ahs, and restarts here are structural features of talk that often accompany delicate situations – situations where, for example, word choice is important, and/or where the action the speaker is trying to pull off has the potential to cause discomfort, embarrassment, offence, etc. These kinds of features can be indicators of interactional ‘trouble’ of some kind. Todd bookends this little conversational detour with an apologetic ‘and again off topic’ at line 971. The point here is to highlight the remarkable work it takes to introduce an idea that does not have clear relevance to the local talk. I mean ‘local’ in terms of immediately prior turns and the current topic, not in terms of other possible scales of ‘local’ such as ‘today’, ‘this project’, ‘this meeting’, ‘this task’, etc. We can quite clearly see that this immediate sense of ‘local’ is the scale at which the participants themselves are trying to display the relevance of their talk (c.f. Schegloff Citation1997). In these three examples neither Todd nor Sandra are working to demonstrate the relevance of their idea to the company, to the project, or to the task, but are working to show its relevance to the current topic of conversation.
At this point we might speculate about the relationship between the two ‘requirements’ of relevance for conversationalists and novelty for brainstormers. Having seen the work required to shift topics, and having seen the attention designers pay to creating demonstrable links to the current topic, we might wonder at the extent to which the normative expectation that conversationalists speak with local relevance is detrimental to the introduction of ideas that are truly novel, or that genuinely break existing mindsets. After all, Sandra's pastry brush seemed to get short shrift. Clearly, the introduction of groundbreaking ideas is unlikely to be a matter of ‘you know, I think …’, or ‘what about …’; rather it is likely to need to find hooks on the current topic from which to justify its appearance in conversation just here. However, one cannot draw the conclusion that design is impoverished by virtue of this conversational commitment to topical relevance. For instance, it is wholly unclear what kind of conversation would transpire in the absence of this conversational ‘requirement’, and there is a significant possibility that interaction without it would fail to be recognisable as a conversation about anything. Furthermore, it is difficult to see how other important aspects of brainstorming, such as building on the ideas of others, would be possible without a shared and accountable expectation that talk be locally relevant.
6 Discussion
While there are many different formats for interaction (e.g. interviews, ceremonies, legal testimonies and presentations are each distinctive in important ways), there are no time outs from social order. Brainstorming sessions may have a character of their own, yet it is a character dependent in essential ways on the same order from which the other arenas of our interactions with each other in daily life are built. In this paper, I have tried to show some of the ways in which designers handle the rules of brainstorming in light of the other orientations they clearly display in interaction, and I have argued these are orientations to the production and maintenance of a normative social order. But what are the lessons for design?
We can and do alter formats of interaction which create different frameworks for participation. Designers are not condemned to accept the format of a meeting or any other social encounter if it does not offer them the flexibility or structure that they require of it. Furthermore, while we have seen that orientations to social order are not suspended on account of the rules of brainstorming, we have also seen that the structure and rules of brainstorming may have had influences on the structures of action within the meeting – notably with the work done by unfinished sentences to invite other-speaker completion of an idea, but also by the fact that participants do occasionally invoke the rules of brainstorming, and of course that they are actually generating ideas throughout. That is, the rules of brainstorming do have an effect on the proceedings, though it may be a milder effect – and one severely modulated on account of social order – than is typically assumed. Methods such as brainstorming are not impotent to structure designers' actions, but nor are they simply and unproblematically ‘obeyed’ or universally oriented to. Much, much more is going on in interaction.
The ‘rules’ of a session or a method (or the stages in a model of the design process) do not easily account for designers' activities, yet they can be (as they were here) among the participants' own accounting devices within the activity itself. Rather than becoming an analyst's resource to account for what designers do (e.g. are they breaking the rules?), they can be an analytic topic for study (Zimmerman Citation1971). This is especially so since, as we have seen here, they are a participant's resource, and since many participants' actions cannot simply be seen as actions in accord with or in breach of brainstorming rules. That is, the method's rules are an occasioned resource on hand for the participants to assign sense, meaning and order to the proceedings in their course (c.f. Garfinkel Citation1967; Zimmerman and Pollner Citation1971; Wieder Citation1974). It is largely in this way that methods come to be of use to designers, to the extent they are deemed by participants to have local relevance for actions to be sanctioned and sanctionable with reference to them. Invoking a rule is one of many means of sustaining social order. Yet although social order is normative (as rules are), it is by no means deterministic. And it is in this space that we may find a practical application for this kind of understanding of design practice. The very identification of designers' normative orientations (e.g. to the local relevance of talk) is one important step towards the creation of formats of interaction that might be able to ‘tamper’ with social order, in similarly mild ways, so as to be more conducive to local design objectives.
Acknowledgements
Max Eckardt provided invaluable help in finding and examining relevant excerpts from the meetings. Trine Heinemann offered incisive comments on an earlier draft, and did much to nurse my sensitivities to conversation analysis. Naturally, I alone am responsible for the errors that remain.
Notes
1. The notion that ‘brainstorming meetings’ exhibit a distinct modification of the structure of ordinary conversation requires empirical study. Not everything that is called a ‘brainstorming meeting’ may share these structural features, and other interactions that have little or nothing to do with brainstorming may also exhibit them. The discussion of Hester and Francis (Citation2001) is important here – conversational structures do not neatly map to our ordinary stock of categories of events.
Related Research Data
References
- Bowers , J. and Pycock , J. . Talking through design: requirements and resistance in cooperative prototyping . Proceedings, Human Factors in Computing Systems CHI '94 . pp. 299 – 305 . Boston, MA : ACM Press .
- Bucciarelli , L. L. 1994 . Designing engineers , Cambridge, MA : MIT Press .
- Button , G. and Sharrock , W. W. 2000 . “ Design by problem-solving ” . In Workplace studies: recovering work practice and informing system design , Edited by: Luff , P. , Hindmarsh , J. and Heath , C. 46 – 67 . Cambridge : Cambridge University Press .
- Coyne , R. and Snodgrass , A. 1993 . Rescuing CAD from rationalism . Design Studies , 14 : 100 – 123 .
- Dennis , A. R. and Valacich , J. S. 1994 . Group, sub-group and nominal group idea generation: new rules for a new media? . Journal of Management , 20 : 723 – 736 .
- Garfinkel , H. 1967 . Studies in ethnomethodology , Englewood Cliffs, NJ : Prentice-Hall .
- Hales , C. and Wallace , K. 1988 . Detailed analysis of an engineering design project . International Journal of Applied Engineering Education , 4 : 289 – 294 .
- Hester , S. and Francis , D. 2001 . “ Is institutional talk a phenomenon? Reflections on ethnomethodology and applied conversation analysis ” . In How to analyse talk in institutional settings: a casebook of methods , Edited by: McHoul , A. and Rapley , M. London : Continuum .
- Jablin , F. M. 1981 . Cultivating imagination: factors that enhance and inhibit creativity in brainstorming groups . Human Communication Research , 7 : 245 – 258 .
- Luff , P. and Heath , C. C. 1993 . “ System use and social organisation: observations on human-computer interaction in an architectural practice ” . In Technology in working order: studies of work, interaction and technology , Edited by: Button , G. 184 – 210 . London : Routledge .
- Matthews , B. 2007 . Locating design phenomena: a methodological excursion . Design Studies , 28 : 369 – 385 .
- McHoul , A. and Rapley , M. 2003 . What can psychological terms actually do? . Journal of Pragmatics , 35 : 507 – 522 .
- Oak , A. this issue. Performing architecture: talking ‘architect’ and ‘client’ into being. CoDesign
- Rittel , H. W.J. and Webber , M. M. 1973 . Dilemmas in a general theory of planning . Policy Sciences , 4 : 155 – 169 .
- Sacks , H. 1995 . Lectures on conversation , Oxford : Blackwell .
- Sacks , H. , Schegloff , E. A. and Jefferson , G. 1974 . A simplest systematics for the organization of turn-taking for conversation . Language , 50 : 696 – 735 .
- Schegloff , E. A. 1996 . “ Turn organization: one intersection of grammar and interaction ” . In Interaction and grammar , Edited by: Ochs , E. , Schegloff , E. A. and Thompson , S. A. 52 – 133 . Cambridge : Cambridge University Press .
- Schegloff , E. A. 1997 . Whose text? Whose context? . Discourse & Society , 8 : 165 – 187 .
- Schön , D. A. 1987 . Educating the reflective practitioner: toward a new design for teaching and learning in the professions , San Francisco, CA : Jossey-Bass .
- Schön , D. A. 1990 . “ The design process ” . In Varieties of thinking: essays from Harvard's Philosophy of Education Research Center , Edited by: Howard , V. A. 110 – 141 . New York : Routledge .
- Simon , H. A. 1973 . The structure of ill-structured problems . Artificial Intelligence , 4 : 181 – 201 .
- Simon , H. A. 1981 . The sciences of the artificial , Cambridge, MA : MIT Press .
- Sutton , R. I. and Hargadon , A. 1996 . Brainstorming groups in context: effectiveness in a product design firm . Administrative Science Quarterly , 41 : 685 – 718 .
- Wieder , D. L. 1974 . Language and social reality: the case of telling the convict code , The Hague : Mouton .
- Winograd , T. 1991 . “ Thinking machines: can there be? Are we? ” . In The Boundaries of humanity: humans, animals, machines , Edited by: Sheehan , J. J. and Sosna , M. Berkeley, CA : University of California Press .
- Zimmerman , D. H. 1971 . “ The practicalities of rule use ” . In Understanding everyday life: toward the reconstruction of sociological knowledge , Edited by: Douglas , J. D. 221 – 238 . London : Routledge and Kegan Paul .
- Zimmerman , D. H. and Pollner , M. 1971 . “ The everyday world as a phenomenon ” . In Understanding everyday life: toward the reconstruction of sociological knowledge , Edited by: Douglas , J. D. 80 – 103 . London : Routledge and Kegan Paul .