Affordances and the Analysis of Technologically Mediated Interaction: A Response to Brian Rappert

Full text

581
Sociology, Vol. 37 No. 3 (2003): 581-589
Affordances and the analysis of technologically
mediated interaction:
A response to Brian Rappert*
IAN HUTCHBY
Brunel University
Brian Rappert takes me to task on a number of grounds in my discussion of the potential of
‘affordances’ as a conceptual resource in the analysis of social interaction involving technology
(Rappert, 2003). These include his claims (a) that my article ‘Technologies, Texts and Affordances’
(Hutchby, 2001a), and the book Conversation and Technology (Hutchby, 2001b) to which Rappert
also addresses many of his remarks,
1
fail adequately to acknowledge that others in the sociology of
technology have developed similar ‘third’ positions between determinism and constructivism, some
of them using the concept of affordances itself; (b) that I unfairly caricature the ‘post-essentialist’
approach exemplified in the work of Grint and Woolgar (1997); and (c) that when it comes down to
the actual analysis of the role of technological artefacts in social life, the concept of affordances does
not in fact offer the kind of fruitful line of inquiry that I suggest it does. Within the space allotted to
me by the Editors, it is impossible for me to respond in detail to each and every one of Rappert’s
claims, and to each and every one of his characterisations of my argument. I will try, however, to
deal with what I believe are the key points at issue between us, and hope along the way to offer some
clarification as to what I was trying to get at in my original article, including both the sociological
significance of that argument and its limits of application.
Let me begin, for the sake of establishing the grounds upon which debate may meaningfully
take place, by briefly recapitulating the key points at stake in ‘Technologies, Texts and Affordances’.
* I would like to thank Susan Speer for her useful comments on an earlier version of this article.
1
Since Rappert freely moves in his remarks between the original article that appeared in Sociology (Hutchby, 2001a) and
the book which appeared the same year (Hutchby, 2001b), I will do the same, while bearing in mind that many readers
may not have read the latter.

582
This is important because Rappert often takes me to task for failing to address issues which he takes
it should be addressed, but which are in fact of a different order to the kinds of questions that I am
arguing need to be addressed. In that article I proposed that the concept of affordances enables ‘a
shift in analytic focus for the sociology of technology: a change in empirical footing’ (Hutchby,
2001a: 450). By that I meant a shift from focusing primarily on questions of representation and
negotiation in understanding the relation between ‘the social’ and ‘the technical’, to a focus on
questions of the use-in-situated-social-interaction of technological devices: specifically, those used in
the mediation of human interpersonal communication (or ‘talk-in-interaction’). I concentrated on the
work of Grint and Woolgar (1997) because I believe that it represents the strongest version of the
anti-determinist/anti-essentialist position in science and technology studies (STS), while equally
demonstrating how difficult researchers in this field find it to get away from the prioritisation of
‘representations’ in their analyses of given technologies. In their ‘post-essentialist’ standpoint, Grint
and Woolgar seek to transcend what they describe as the continuing adherence to a ‘dualism between
“technology” and “the social”’ (1997: 21) pervading not only so-called technological determinist but
also social constructionist work in STS – including many of the examples cited approvingly by
Rappert in the first section of his article. This, I maintain, is the main strength of the technology as
text metaphor: on the one hand it stands at the furthest remove from determinism; while on the other,
it identifies a flaw in the reasoning of key constructionist schools of thought, which Grint and
Woolgar (1997: 37) refer to as ‘residual technicism’. The text metaphor presents STS with the
specific challenge of attempting to purge all rogue references to the ‘inherent properties’ of
technologies, and Grint and Woolgar advocate the policy of ‘analytical scepticism’ in which every
claim, including those of the author, should be opened up to the infinite regress of questions
regarding its own epistemological basis.
However, I argued that the technology as text metaphor sustains by its very nature the focus
of the sociological gaze on questions of representation, interpretation and negotiation which, for me,
inevitably deflect attention from the empirical question of embodied human practices in real time
situated interaction involving technologies. This is clear in Grint and Woolgar’s (1997) main
rhetorical strategy, which is to show, first, that any artefact lends itself to more than one possible
interpretation, then to take selected interpretations and show the degree to which they rely upon some
conception of the technology’s ‘inherent’ characteristics, and finally to state that the only way of
avoiding such ‘essentialism’ is to focus on analysing the discursive practices through which one
interpretation wins out over others. The best example of this is in the final chapter of Grint and
Woolgar’s (1997) book, in which they introduce a new metaphor – the onion model – to demonstrate
that ‘every successive attempt to reach a final, uninterpretable “effect”...from a determinate
“cause”...can, with enough stubbornness, counter-intuition and effort, be construed as (yet another)

583
social construction’ (Edwards et al. 1995: 37).
My argument was that despite what the authors might state, the logical implication of the
technology as text metaphor is that ‘what [technologies] are is a matter for negotiation and
persuasive rhetoric’ (Hutchby, 2001b: 27); and the upshot of this is that technological phenomena are
considered as tabulae rasa. Rappert argues that this is a misrepresentation, and instead prefers a
weaker interpretation; namely that post-essentialists merely ‘try to be aware of the premises and
contingencies underlying definitive characterisations’ (Rappert, 2003: 571). Yet when we examine
what Grint and Woolgar themselves say, we find that they state something much closer to my own
version in the following crucial passage: ‘In disassociating the upshot of reading and interpretation
from any notion of the inherent quality of the text...we do not mean to suggest that any reading is
possible (let alone that all readings are equally possible), although in principle this is the case’ (Grint
and Woolgar, 1997: 72-3, emphasis added). Clearly, the authors did not need to add that final
(italicised) clause; and if they had not, then perhaps Rappert’s account of their position would have
been acceptable. The fact that they did raises the following logical conundrum: either any reading is
possible (even if only ‘in principle’) – a position which Rappert admits is absurd – or not all readings
are, in fact, possible (despite what the authors say) – in which case we are faced with the question,
what is it that constrains the possible readings that can be made of a given technology? The answer I
proposed is that these constraints are best understood as rooted in the array of affordances that a
technology makes available, since what a technology affords provides ‘the very conditions of
possibility for competing accounts to be sensibly made’ (Hutchby, 2001a: 450). By extension, the
affordances of technologies also furnish the conditions of possibility for certain actions – and
actions-in-interaction – to be carried out via, with, or in relation to that technology.
It is this that is at the heart of the brief case study at the end of ‘Technologies, Texts and
Affordances’ (Hutchby, 2001a: 451-453). In that passage I offered an alternative account of some
data discussed by Grint and Woolgar (1997; see also Woolgar, 1991) in which, in mundane terms, a
participant in a usability trial for a new computer model experiences difficulties carrying out the task
of plugging a printer into the system because, as it turns out, she has unwittingly been given a printer
cable which was not designed to fit this particular model. Rappert defends Grint and Woolgar’s
account of these events in terms of ‘uncertainties experienced at the time where questions over who
had the authority to speak for the computer system loomed large’ and caricatures my argument as a
suggestion that ‘the proper way to interpret the situation was that the plug did not afford plugging in’
(Rappert, 2003: 575). This, however, is to seriously misrepresent the argument I was in fact making,
which was to shift the focus of attention such that the problem could be seen in a wholly different
way; one which showed that ‘the relationship between user and technology is bounded not so much
by a politics of speakership and representation, but by ordinary practices interfaced with material

584
enablements and constraints’ (Hutchby, 2001a: 453).
My central aim in ‘Technologies, Texts and Affordances’ was to establish the significance of
a sustained and analytically rigorous program based on investigation of the nature of ordinary actions
in the context of such material enablements and constraints. Drawing in part on the work of others
who are, and for some time have been, developing such an empirical program of work, in
Conversation and Technology I argued that the concept of affordances offers an important yet
hitherto unacknowledged means of synthesising a disparate body of empirical investigations of
‘technologies in situated social interaction’, because it is a means by which both the materiality of
technologies themselves and the observable orientations of technology users can simultaneously be
taken into account.
As Rappert correctly infers, this project draws heavily on the ethnomethodologically inspired
program of conversation analysis (CA).
2
CA is concerned with the description and analysis of the
normative structures underpinning the production of talk in ordinary situations of social interaction:
not only ‘casual’ conversations, but any form of concerted activity where talk-in-interaction
constitutes a primary vehicle whereby the work of an encounter gets done. Among CA’s hallmarks is
the notion that the normative structures underpinning interaction are oriented to by participants
themselves, and thus are available to analysis by means of close observation of participants’
displayed orientations to what, at any given moment, they take it their interaction is ‘about’. This can
range from doing ‘asking and answering a question regarding the weather’ (e.g. Pomerantz, 1984) to
doing ‘conducting cross-examination of a witness in a trial for rape conducted under the auspices of
the adversarial legal system’ (e.g. Drew, 1992).
How does this approach inform a sociological understanding of technologies and their role in
the organisation of interaction? A good proportion of early conversation analytic studies were carried
out using data derived from technologically mediated interaction: talk on the telephone. But while
telephone conversation brings particular methodological benefits for the conversation analytic
enterprise, since the technology filters out interactional resources such as gesture and gaze, giving
the analyst of an audio recording access to the same set of turn-taking practices as the participants
themselves,
3
it also raises its own new analytical question: since the terms of interaction are altered
(e.g. there is no mutual visual access), the organisation of interaction itself may manifest systematic
alterations. If so, what is the nature of those alterations and how are they related to the technological
2
The seminal studies in conversation analysis were conducted by Harvey Sacks and are published in Sacks (1992). Other
important collections of studies include Atkinson and Heritage (1984), Boden and Zimmerman (1991) and Drew and
Heritage (1992). Hutchby and Wooffitt (1998) is a full-length introduction to the field.
3
Conversation analysts have also analysed the uses of gesture, gaze and other elements in the organisation of talk-in-
interaction using video recording. See, for example, Goodwin (1981).

585
medium?
4
It is not clear to what degree CA researchers actually seriously addressed that question, which
involves considering the relationship between technology, mediation and the practices of talk-in-
interaction. The general stance in CA based on data derived from telephone conversation has been
that the technology should be treated not ‘as itself the interest of the first order, but...as a device
through which are refracted other phenomena’ (Schegloff, 2002a: 290). This results in a somewhat
ambivalent position in which, as Schegloff (2002a: 293) puts it, ‘the point of [telephone-based]
studies is not conversation on the telephone per se...but the telephone as another tool for the analysis
of talk-in-interaction, as especially constituted in this technologically shaped context’ (Schegloff,
2002a: 293). The status of the ‘technologically shaped context’ itself remains unclear, however.
While Schegloff (2002b: 324) acknowledges that ‘this work...did spawn a line of research for which
the technology involved was central – a whole genre preoccupied with telephone conversation’ he
equally maintains that the technology ‘served to spur the development of a whole modality of
research to which it was fundamentally peripheral, but to which it provided a contextually varied
glimpse of its basic data’. Yet in discussing the advent of mobile telephony, Schegloff (2002a: 297)
is nevertheless moved to ask: ‘how should we understand cell phone use: is it like any other phone
use, or do the new technological affordances modify the terms under which such conversations are
initiated and conducted?’.
It is striking how similar this question is to the one that drove my own work in Conversation
and Technology (Hutchby, 2001b), where I sought to bring the concept of affordances into play in
order to make more explicit, and more systematic, the terms on which we might explore the
relationship between technological mediation and the course of social interaction. The technologies
to which this applies are not restricted to the telephone: contrary to Rappert’s assertion that
Conversation and Technology is concerned only with ‘a highly limited set of issues regarding the
place of telephones in society’ (Rappert, 2003: 574), the book contains studies of technologies other
than, and very different from the telephone, such as video-linked computer terminals, ‘expert
systems’, ‘intelligent’ interfaces, computers with ‘conversational’ capabilities, and real time
exchange via internet connections. Collectively, I referred to these as ‘technologies for
communication’: the subset of technological artefacts whose main use is in the mediation of
interpersonal communication. The key project undertaken in Conversation and Technology was to
investigate whether there may be specific forms of interaction that have grown up around such
technologies for communication; and its argument centred upon ‘a complex interplay between the
4
Other researchers, not working within the conversation analytic tradition, have asked a similar question, though their
answers differ markedly from mine. See, for instance, the work of Rutter (1989), which I discuss in Conversation and
Technology (Hutchby, 2001b: 86-88).

586
normative structures of conversational interaction and the communicative affordances offered by
different forms of technology’ (Hutchby, 2001b: 13, emphasis in original).
Rappert fails to appreciate the complexity of this interplay, largely because he seems
concerned to take me to task for not addressing the issues which he takes it should be addressed in
the sociology of technology. At various places, he asserts that a focus on the affordances of
technologies produces insights that are interesting but only in a ‘limited way’. The implication is that
my approach ultimately fails to engage with the ‘real issues’ at stake in the sociology of technology.
These issues include, for example, ‘the relation between technology, politics and truth’ (Rappert,
2003: 572), or ‘the relation between technical decisions, representations and uses of technology’
(Rappert, 2003: 573). In other words, Rappert prioritises what he takes to be the known big issues in
the sociology of technology, over what he takes to be the limited issues to which my approach is
addressed. This is an effective rhetorical strategy, one that pervades social science theorising and is
deeply bound up with sociology’s own self-justification. As Sacks (1984: 22) once remarked,
theories in the social sciences have historically been based on a distinction between ‘what are in the
first instance known to be “big issues”, and...those which are terribly mundane, occasional, local and
the like’, where only the former are considered proper subjects for serious research. More or less
explicit in Rappert’s deployment of this distinction is the equally common view that ‘the search for
good problems by reference to known big issues will have large-scale, massive institutions as the
apparatus by which order is generated and by a study of which order will be found’ (Sacks, 1984:
22): institutions such as the research and development wings of major technical corporations, or the
policy-making bodies attached to national government. But if, as Sacks went on to suggest, ‘we
figure or guess or decide that whatever humans do, they are just another animal after all, maybe more
complicated than others but perhaps not noticeably so, then whatever humans do can be examined to
discover some way they do it.... That is, we may alternatively take it that there is order at all points’
(1984: 22, emphasis added). This suggests an alternate research agenda in which ‘whatever humans
do’ is taken seriously whether or not the activities being observed and described can be seen and
interpreted in terms of any ‘known big issues’.
The upshot is that for me, the relationship between technologies and social processes is
interesting at the empirical level of the pursuit of concrete courses of interaction. The social
processes in question are those of talk-in-interaction, and their significance is in the manifestation of
interlocutors’ orientations to the constraints and enablements of material technological devices. In
other words, it is in the interface between the organised practices of human communication and a
given technology’s array of communicative affordances that we can find the significance, for its
users in their real-world courses of interaction, of the technology as a worldly artefact.
Thus, when Rappert states that my argument involves treating technologies ‘as a stable,

587
predictable and functioning part of the practices of everyday life’ (Rappert, 2003: 570), he is right,
but only partly so. It is not I, as a sociologist, who advocates this approach to technological artefacts.
What I in fact say is that ‘people rely on the very materiality of the telephone, its stability as an
object in the world, and its predictability as a functioning medium of communication, in order to
carry out the tasks for which they seek to use it’ (Hutchby, 2001b: 198, emphasis added). Rappert’s
omission is significant. It indicates that he is more interested in the nature of sociological claims
about technology than in ordinary societal members’ uses of technology. For instance, he contends
that the concept of affordance ‘is not much help in elaborating the implications of technology or the
interrelatedness of technology and the social’ (2003: 574). But it seems clear that ordinary actors,
when they encounter and try to use a device, are not concerned with such things either. What they are
concerned with is what the technology can and cannot do, and what they, in their situated courses of
action, can and cannot do with it.
5
Rappert also suggests that sociologists should perhaps try to
‘determine which descriptions of the affordances of technology are actually correct and simply
consider others as misrepresentations’ (2003: 577). But surely it is the users of technologies, ordinary
societal members, whose activities in determining the affordances of those technologies should be of
interest. In this, the question should rather be: ‘what are participants using a technology to do, and
how are they doing that: how are they orienting to the technology (or more strictly, to what it enables
and constrains in terms of ongoing courses of action – its affordances) as a factor in the situated
activities in which they are engaged?’ This is what I mean by a ‘change in empirical footing’. Like
many examples of ‘reflexive’ sociology, Rappert’s critique ultimately becomes too fixated on the
activities and priorities of members of the category ‘sociologists’. I believe, instead, in prioritising
the activities of members of the category ‘ordinary social actors’.
It is in the nature of these things that there are many points in Rappert’s article to which I
would like to respond but, given space constraints, have been unable to. What I have tried to do is
provide, for the interested reader, a clarification of my own argument, which I believe is
misrepresented at numerous points in Rappert’s piece. In the process it should have become clear
that Rappert’s challenge, in the early part of his article, to the originality of my argument is
misplaced. While it is true that others in the sociology of technology have sought so-called ‘third
ways’ between determinism and constructivism, mostly by means of realist epistemologies and
sometimes utilising versions of the affordances concept itself, the studies he cites are nevertheless
rooted in an analytic framework that prioritises technologies as the object of attention. My analytic
framework instead prioritises actions: more specifically, actions in the course of mundane interaction
involving some degree of technological mediation. My work has not been concerned with the nature
5
See, for example, Suchman’s (1987) influential work on situated action in the context of expert systems design.

588
of technologies per se (be that ‘essential’, ‘socially shaped’, ‘textual’ or whatever) but with the
nature of human interaction with, or via, certain technological devices. Thus, the primary interest is
neither in conversation nor in technology, but in conversation and technology: in those observable
events in the social world where it appears that a technology – or more strictly its affordances –
comes to play a role in the exchange of turns at talk, in the structures of those turns, and in the
actions accomplished by those turns.
REFERENCES
Atkinson, J.M. and Heritage, J. (eds.) 1984. Structures of Social Action: Studies in Conversation
Analysis. Cambridge: Cambridge University Press.
Boden, D. and Zimmerman, D. (eds.) 1991. Talk and Social Structure. Cambridge: Polity.
Drew, P. 1992. ‘Contested evidence in courtroom cross-examination: The case of a trial for rape.’ In
P. Drew and J. Heritage (eds.), Talk At Work. Cambridge: Cambridge University Press.
Drew, P. and Heritage, J. (eds.) 1992. Talk At Work. Cambridge: Cambridge University Press.
Edwards, D., Ashmore, M. and Potter, J. 1995. ‘Death and furniture: The rhetoric, politics and
theology of bottom line arguments against relativism.’ History of the Human Sciences, 8: 25-
49.
Goodwin, C. 1981. Conversational Organisation. New York: Academic Press.
Grint, K. and Woolgar, S. 1997. The Machine At Work. Cambridge: Polity.
Hutchby, I. 2001a. ‘Technologies, texts and affordances.’ Sociology, 35: 441-456.
Hutchby, I. 2001b. Conversation and Technology: From the Telephone to the Internet. Cambridge:
Polity.
Hutchby, I. and Wooffitt, R. 1998. Conversation Analysis. Cambridge: Polity.
Pomerantz, A. 1984. ‘Agreeing and disagreeing with assessments: Some features of
preferred/dispreferred turn shapes.’ In J.M. Atkinson and J. Heritage (eds.), Structures of
Social Action: Studies in Conversation Analysis. Cambridge: Cambridge University Press.
Rappert, B. 2003. ‘Technologies, texts and possibilities: A reply to Hutchby.’ Sociology, 37: 565-
580.
Rutter, D.R. 1990. ‘The role of cuelessness in social interaction: An examination of teaching by
telephone.’ In D. Roger and P. Bull (eds.), Conversation. Clevedon: Multilingual Matters.
Sacks, H. 1984. ‘Notes on methodology.’ In J.M. Atkinson and J. Heritage (eds.), Structures of
Social Action: Studies in Conversation Analysis. Cambridge: Cambridge University Press.
Sacks, H. 1992. Lectures on Conversation. Oxford: Blackwell.
Schegloff, E.A. 2002a ‘Beginnings in the telephone.’ In J.E. Katz and M. Aakhus (eds.), Perpetual
Contact. Cambridge: Cambridge University Press.

589
Schegloff, E.A. 2002b. ‘On opening sequencing.’ In J.E. Katz and M. Aakhus (eds.), Perpetual
Contact. Cambridge: Cambridge University Press.
Suchman, L. 1987. Plans and Situated Actions. Cambridge: Cambridge University Press.
Woolgar, S. 1991. ‘Configuring the user: The case of usability trials.’ In J. Law (ed.) A Sociology of
Monsters: Essays on Power, Technology and Domination. London: Routledge.

590
APPENDIX (Not included in the original publication)
On Saturday, December 2nd 2000, The Guardian newspaper published the following story:
MAN ARMED WITH TWIGLET JAILED
Anthony Newton, 57, was jailed for four years at Exeter crown court after admitting holding
up an off-licence in Torbay, Devon, by pushing a Twiglet into the cashier’s ribs and telling
her to hand over the money.
A postgraduate student circulated this snippet by e-mail to staff in my department with a playful
query as to what kind of sociological case might be argued about it. At the time, I was working on
the final stages of the texts to which Brian Rappert responds in his article in this issue: namely, the
polemical article ‘Technologies, Texts and Affordances’ which appeared in this journal (Hutchby,
2001a) and the book, Conversation and Technology (Hutchby, 2001b) which empirically fleshes out
the case made in ‘Technologies, Texts and Affordances’.
My contribution to the e-mail discussion centred, not suprisingly perhaps, around the
suggestion that the news item indicated something about the affordances of Twiglets: i.e., that in
virtue of certain aspects of their materiality they could be used, in certain circumstances, to simulate
the kind of weapons (guns, knives) commonly used in off-licence hold-ups. Clearly, this particular
instantiation of (one of) the affordances of Twiglets relies for its success on aspects of the context in
which the affordance is being brought into play. The participants are encountering one another in an
off-licence, one as an employee and the other as, at least putatively, a customer. Off-licences are one
context in which ‘being held up’ is what could be described as an omnirelevant action category. That
is, employees, finding the putative customer thrusting a sharp object into their ribs, have immediately
(and perhaps preferredly) open to them the interpretation that ‘this is a hold-up’. Thus, although the
relationship between the Twiglet and the gun or knife is only synechdocal, and although the Twiglet
is not the only object that affords such a synechdoche, in this particular context ‘being used to
simulate a gun or knife’ is one affordance of Twiglets that becomes interactionally available.
In response, one colleague wondered whether this meant that the affordances of Twiglets
were defined a priori, with ‘able to be used in a hold-up’, therefore, as one of them. This is similar to
Rappert’s own suggestion that my work on technologies and affordances is concerned with
‘establishing the definitive properties of technology in a manner that links expectations, uses and
representations’ (Rappert, 2003: 569, my emphasis).

591
I profoundly disagree with this interpretation; and indeed a careful reading of the arguments
presented in Hutchby (2001a) and (2001b) will reveal that nowhere in either text do I make any
claims about establishing ‘definitive properties’ of technologies. The affordances of things do not
need to be defined, or indeed known about a priori in order for them to be brought into play in some
course of social action, nor for ‘new’ ones to be discovered and instantiated in the course of social
action. That is to confuse the issue, for the concept of affordances (contrary to what Rappert, along
with my colleague, seems to believe) is not so much about the properties of objects themselves, as
about what agents (be they human or otherwise) are able to do with objects, in the course of agentic
action projects carried out in relation to particular interactional environments. In Conversation and
Technology I therefore stress the extent to which affordances are both functional and relational.