Within the literature on organizations there has been a movement away from mechanistic, individualistic and cognitivist approaches to organizational behavior (Weick, 1979; Schon, 1983; Engstrom, 1987; Blackler, 1995). These developments often stress the limitations of such approaches in helping organizational practitioners to cope with the often perplexing, uncertain and ambiguous conditions of organizational life— conditions that seem particularly acute in contemporary capitalism—and their failure to fully appreciate the sociocultural dimensions of activity. Approaches to skill and skilled activity have not been immune from these critiques and developments. “Mainstream” cognitive and social psychological approaches to skill have been criticized for being too mechanistic, for ignoring the social and cultural aspects of skilled activity, for being unable to explain fully how people cope with novel variation in a context, and for being insensitive to the way in which the meanings of skill are contextually situated (Scribner, 1984; Holman, 1999). Furthermore, mainstream approaches appear to assume that actors need a fairly clear understanding of their context and the skill they are using before skilled activity can take place (Hargie, 1997). This seems to contrast with studies that indicate that such understandings are not necessary, that too concrete an understanding may hinder skilled activity, and with the fact that human action, and skilled activity in particular, may be best characterized as “the transition from vagueness to clarity [and] from uncertainty to decision” (Bereiter, 1991: 13).

The dialogical approach to skill and skilled activity, which draws on ethnomethodology (Garfinkel, 1967; Heritage, 1984), activity theory (Vygotsky, 1978), and social constructionism (Billig, 1987; Shotter, 1995), was developed in response to the problems and limitations of mainstream approaches (Holman, 1997, 1999). It suggested that the dialogical approach is better able to account for the social, contextually situated, creative, contingent and responsive nature of much skilled activity. In particular, the dialogical approach highlights the narrative and argumentative features of skill, how these are used to make sense of uncertainty, and how they are used to develop valued certainties and social and moral orders. It also views skilled activity as a joint activity, and seeks to understand skill and skilled activity within its wider social and cultural context.


Skill can be broadly defined as a personal ability that has the potential to enable a person to act effectively. Skilled activity is the social expression of skill use. Skills are understood to be the product of internalized interpersonal action. Being a result of internalized interpersonal activity, skills reflect the form of the interaction from which they are derived. In particular, skills reflect the dialogs from which they are derived and retain a sense of the forms and functions of discourse used in dialog.

At a general level, discourse can be understood as having narrative and non-narrative forms (Weick and Browning, 1986; Georgakopoulou and Goutsos, 1997). Narratives are concerned with the temporal ordering of ideas and actions, and focus on their sequential patterning, their duration and pace, their context and the role of actors. Non-narratives are concerned with non-temporal patterns such as the relationships between propositions, argumentation, descriptions, ideas and evaluations. Discourse can also be understood as having three functions (Halliday, 1985): an ideational function, e.g. the communication of ideas, metaphors; an interpersonal function, the achievement of social action through talk, e.g. arguing, questioning, promising; and a textual function, the precise organization of language into a linguistic event. The dialogical approach tends to focus on the interpersonal functions of skill.

The narrative features of skill

If narrative involves the temporal patterning of events and actors over time, then a skill must capture this patterning and enable its re-enactment over time. Each skill has a range of narrative features and these affect how a skill structures the unfolding of activity over time (Bahktin, 1981). The important narrative features of a skill are knowledge of:

  • Content—what and how an action should be done, and the likely effects of an action.

  • Ordering—in what order action should occur.

  • Timing/temporality—the pace, duration and synchronization of action.

  • Context appropriateness—where an action can occur.

  • Actor eligibility—who is allowed to do what and to whom.

Narrative features can be seen in relatively “simple” skills such as asking questions, making requests and offering assessments, and in more complex and temporally extended skills such as telephone interviews, arguments and group decision making (Hootkoop, 1987; Pentland and Rueter, 1994; Poole and Holmes, 1995). For example, studies on group decision making have demonstrated that in order for its effective and efficient enactment, each group member needs to have some knowledge of the general process of decision making (content), the interpersonal actions required to make the process possible (content, context appropriateness), the possible ordering of interpersonal actions, and the timing of activity both at a micro level (e.g. pauses in speech, how long to speak for) and at a macro level (e.g. how long to spend discussing a solution) (McGrath, 1991).

Non-narrative features of skill

While there are a number of non-narrative features of skill, perhaps the most important is a skill’s argumentative structure. Using Toulmin’s (1958) analysis of arguments, it is suggested that there are three argumentative features of a skill: claim, ground and warrant. Claims are the statements put forward for judgment, the grounds are the data used to support a claim, and the warrant connects the grounds to the claim (Fletcher and Huff, 1990). When a person uses a skill, each of the narrative features deployed makes a claim about being an appropriate means of structuring action, and this, often implicitly, will be supported by the grounds and the warrant that justify it.

For example, in a group that is solving a problem and has come up with a number of possible solutions, the statement by a group member “Time’s passing, do you think we should choose one now?” can be read as a claim about the need to move towards a more definite solution. The argumentative features of this act might therefore be:

Grounds: (Given that) We have been discussing this issue for five hours.
Warrant; (And since) Our objective is to develop a solution by the end of the day.
Claim: (I assert that) We need to start to choose one solution from the many.

Skilled activity

When a person develops and shapes skilled activity, one of their basic tasks is to generate and create a visible sense of moral and social order from the ambiguity and vagueness that exists in any one context—a move from uncertainty to certainty (Garfinkel, 1967). Skilled activity proceeds by the person using a range of narrative and non-narrative resources (which are part of their skills) with which to create a dialog with themselves, others and artefacts. Early on in the construction of skilled activity, the dialogical process is more open and less constrained. As dialog progresses, the person starts to make sense of their context by constructing more certain and concrete meanings, and feelings may emerge within the person that a sense of moral order has been achieved.

Developing skilled activity also involves the person assuming a set of shared common-sense “background expectancies” with which all actors are motivated to comply (Garfinkel, 1967). Important background expectancies include assumptions about actions being appropriate to a context and that a reasonable and rational person ought to act in a understandable and legitimate manner. Assumptions about the subjective state of other people’s minds (e.g. perspectives, intentions, desires and motivations) also need to be made if a person is to make their actions understandable and legitimate (Schutz, 1962). A further feature of skilled activity is that when a person forms their actions, they do so with an awareness that each may be accepted or contested and, if contested, that they may need to reform that action (Bahktin, 1981). As such, a person may often need to persuade others that their actions are legitimate. The overall “ebb and flow” of skilled activity can be characterized as rhetorical, persuasive and argumentative (Shotter, 1993). Another feature of skilled activity is that it is responsive to contextual contingencies and is thus a creative process.

Skilled activity shares many of the features associated with artful activity (e.g. the creation of a moral order, use of background expectancies, rhetorical nature of activity). Yet, while it has many similarities, it is different in one crucial respect. One of the main distinguishing features between artful and skillful activity is that skillful activity is consensually agreed as doing something better or more effectively than another; and what is seen to be better or more effective is always based on value judgments related to the sociohistorical context. Therefore, unlike artful activity, skilled activity is not just about the creation of certainty from uncertainty, nor is it just about the creation of a moral order. Rather, skilled activity is about the creation of more valued certainties from uncertainty, the creation of more valued actions and outcomes, and the creation of a highly valued moral order. This is not to say that there is no value in artful activity, it is merely that skilled activity is valued more. Skilled activity can be seen as existing on the back of artful activity.

A feature of skilled activity is thus the deployment of valued resources (i.e. skills) in a legitimate and appropriate manner. It is the argumentative structure of the skill that provides the justification for whether a resource is considered to be valued or not A skill will have an argumentative structure that is believed to be a more valid and legitimate means of structuring action than an “ordinary” resource (the difference does not lie in a skill having an argumentative structure that an ordinary resource does not). The argumentative structure of a skill provides its persuasive “power” and this persuasive power can be implied or be made explicit.4 A person using an “ordinary” resource, while able to generate activity in an artful manner, may have difficulty in generating highly valued certainties and moral and social orders. The person using a skill may have less difficulty in generating valued certainties and orders.

An illustration

The following example will be used to illuminate the narrative and nonnarrative features of skill and skilled activity, and in particular its argumentative aspects. This example, taken from a paper by Firth (1995), shows a telephone conversation between Gupta, a cheese importer in Saudi Arabia, and Hans, a cheese seller in Denmark. The conversation is about the price of Hans’s cheese and has been initiated by Hans in response to a telex from Gupta stating that Hans’s prices were higher than those of other cheese importers and requesting a reduction in price.

Hans opens the account by pursuing the non-acceptance of the prices and Gupta responds by explaining why he rejected this offer. At this point Gupta is merely restating what was in the earlier telex. Both parties can therefore be understood as stating their positions with respect to the cheese prices. The passage ends with a pause indicating that Gupta has finished his turn and that it is now Hans’s turn to talk.

Hans responds by probing and challenging Gupta’s assertion about the types of cheese for which he has prices. Although Gupta does not respond directly to this probe, it opens up his account of the reasons for the rejection, a competitor’s cheaper offer. Gupta is starting to lay his negotiating cards on the table and “always we believe your prices” has an ultimatum quality about it suggesting that if Hans does not reduce his prices Gupta will buy elsewhere. Hans then specifies his problem, the reduction in the EEC subsidy, and suggests that he will reduce his prices by this amount.

Gupta rejects this new offer as his alternative offer is still below what Hans has just offered. What follows is an extended period of joint problem-solving activity in which Hans further questions the reasons for Gupta’s rejection. These questions relate to whether transport costs are included in the competitor’s price (they are), the color of the cheese quoted on, and the range of weights that the cheese will come in. In response to each of these probes Gupta offers an account that supports the reasons for the rejection. From this series of questions, Hans appears to have a “challenge trajectory” that attempts to maneuver Gupta into a vulnerable negotiating position (Firth, 1995). It is only after this extensive challenge to the reasons for Gupta’s rejection that Gupta finally reveals the price quoted by the competitor, 3290/ton, whereas Hans is selling at 3404/ton—a difference of about 100.

This sequence is marked by Hans indicating to Gupta that he is calculating the price of the cheese. Hans does this by producing “intendedly hearable self-talk”, which in this case is almost inaudible and which can be used to signal a change in participation status, e.g. from talk to calculation (Goffman, 1981). This appears to be accepted by Gupta as he does not interrupt the long periods of silence that occur. Hans also seems to concede to Gupta’s reasons for the non-acceptance of Han’s cheddar cheese prices, and actually shares in the production of Gupta’s reasoning. Hans makes a new offer that matches his competitor’s offer on cheddar cheese. The negotiation continues, with Hans offering $20 a ton off feta cheese. The cheddar and feta prices are accepted by Gupta telling Hans to send a telex to confirm his offer. The conversation then moves on to another topic.

The content narrative features (i.e. the interpersonal function of the acts) include the setting out of positions, probes, questioning, challenge trajectories, making offers and concessions, and defending positions. Such interpersonal acts are typical of negotiations (Scott, 1989; Mulholland, 1991). The ordering of the negotiation can be seen in the way that it proceeds from each participant stating their position, through both sides defending their position and questioning that of the other, to concessions being made and a resolution achieved. Again, this is a fairly typical ordering of a negotiation. Another example of an act of ordering is the delay or cushioning that occurs before the explanation of the rejection. This has been shown to be important if rejections are to be seen as legitimate (Davidson, 1984; Heritage, 1984).

It is difficult to make claims about the overall timing of the negotiation, apart from the possibility that there is an expectation that the issue will be resolved quickly. At a more micro level, the timing of pauses is used to signify that the speaker has ended his turn in a sequence (e.g. question-answer). Also there is a pause of 3 seconds just after Hans’s offer. Firth suggests that Gupta does not respond immediately, but tries to use a pause to elicit a further concession. He does this again after an ambiguous “huh hmm.” When a further concession is not forthcoming, Gupta accepts the offer.

With regard to the context appropriateness of the acts, perhaps the most striking example is Hans’s use of “intendedly hearable self-talk” to signify a change in participation status. This shows Hans’s sensitivity to communication by telephone. The “actor eligibility” aspect of skill is evident in the way that each expects the other to act as an organizationally responsible agent, and not, for example, as a friend who will cut prices purely because he likes them. This role clearly constrains what they can and cannot do.

Each of the narrative resources deployed by Hans and Gupta makes a claim about being a legitimate and appropriate means of structuring the negotiation. In most cases the claim remains implicit in the act, as do the grounds and warrants. However, total agreement on the reasoning behind an act may not be needed if it is to be regarded as legitimate and for action to be successfully coordinated (Turner, 1994). It is quite possible that, while Hans may accept the act as valid, he may do so for very different reasons. Such a process clearly involves making assumptions about the other person’s state of mind, their interests, and the intended effect of another’s act (Bonaiuto and Fasulo, 1997).

Although not present in Hans and Gupta’s dialog, the argumentative structure of an act can also be used reactively to justify a contested claim. If, for example, Gupta had “misread” Hans’s self-talk, he may have asked what Hans was doing. Hans may then have had to indicate what he was doing and the reasons for his self-talk. Indeed, when beliefs and actions are challenged, justifications and explanations are common, almost morally required, actions (Antaki, 1994). In the passages above, the challenges that were made were done so with regard to ideational differences the extent of a concession) rather than the validity of the acts used whether it was valid to make a concession at that point). There are, however, several points where the legitimacy and appropriateness of an act might have been questioned. The first occurs when Gupta ignores Hans’s question. Another occurs when Hans only reduces his price by the cost of the EEC subsidy. This might have been interpreted as an act of dishonesty, as Gupta knows that other importers are selling for $100 less and that Hans could also probably sell his cheese for less. Instead, Hans’s initial offer appears to be seen by Gupta as a legitimate ploy in the negotiation process.

Hans and Gupta’s conversation can be viewed as a dialogical process in which they use a range of resources with which to coauthor a negotiation narrative. At the beginning of the interaction, neither party, particularly Gupta, can know for sure whether the other is willing to negotiate. Hans and Gupta use their knowledge of the narrative content and ordering of negotiation to make opening statements that clarify any ambiguity about the possible need to frame the following conversation as a negotiation, to “test” their implicit claim as to the appropriateness of this move and to discover each other’s perspective. Hans’s opening question probes the non-acceptance of the cheese prices in Gupta’s telex. Gupta justifies this non-acceptance but, instead of responding to Hans’s next question directly, he opens his bargaining position by stating that he has received a cheaper offer elsewhere. As a result of these actions, and the fact that a challenge to their appropriateness is not made, it becomes quickly apparent to both parties that they are both are willing to negotiate further (Firth, 1995). In this opening phase the context is not pre-given but has to be created. Each act reflexively creates the context, generates a greater certainty and dispels any ambiguity about the nature and purpose of the conversation. The opening of the negotiation is a further reminder that skills are always used “for another first time” (Garfinkel, 1967: 9).

After this opening, the conversation proceeds with Hans and Gupta using a variety of narrative and non-narrative resources to shape the negotiation, although much of the activity now focuses on accounting for the price of cheese and working towards an amicable resolution. The resources are used on a turn-by-turn basis and in a way that is contextually situated and responsive to contingent demands. For example, Gupta must respond appropriately to Hans’s questions and probes, and Hans must respond to Gupta’s different reasons for the rejection of his prices— questions and reasons that cannot be fully known in advance.

There is also a strategic aspect to this process, with each trying position the other into a vulnerable position. The smooth progress of the negotiation not only attests to the skill of both parties, but suggests that Hans and Gupta perceived each other’s acts to be legitimate and deployed in a valued manner. The argumentative aspects of a skill were deployed fairly infrequently (e.g. Hans stating that he accepts Gupta’s arguments and that he will make a concession) and much of the argumentative structure was left implicit. It can be suggested from this that skilled activity may be “easier” when both parties construct a narrative in which they perceive intentions and understandings to be similar.

However, action does not always proceed so smoothly. If, for example, one of Hans’s colleagues (whom we will call Peter) had a much more aggressive approach to negotiation, then negotiations between Peter and Gupta could be characterized by each party being more inclined to question the legitimacy and appropriateness of each other’s acts, e.g. Gupta: “That’s a fairly blunt question, Peter.” Gupta may perceive Peter as using an ordinary and less valued resource. In these instances, where differences in understanding and approach are greater, more narrative features will need to be deployed in order to create valued social orders. It might also be expected that a greater argumentative effort would be needed to persuade others of the appropriateness of one’s actions. This will clearly make the non-narrative features of the skill much more explicit and visible. Thus, if Gupta did challenge the bluntness of Peter’s questioning, Gupta might have said why he found it blunt, and Peter might have had to repair the conversation by apologizing and/or explaining why he thought it was appropriate.

Finally, there are three important valued outcomes to the negotiation. The first, and the most visible, pertains to the development of a resolution to their ideational differences about the price of cheese. The main resources used here are the explanations and reasons presented to each other to support and/or question each other’s position on cheese prices. The second outcome relates to the development of an appropriate and legitimate act of negotiation, i.e. the development of a social and moral order. The resources used here are the narrative and non-narrative features of a “negotiation skill.”

The third outcome relates to the maintenance and confirmation of Hans and Gupta’s identity as proficient organizational members. Firth (1995) suggests that the need to produce actions that are organizationally defensible is an overriding concern of both actors in this negotiation. For example, when Gupta refers to an “us,” it is clear that he needs to arrive at a cost that is justifiable to the “organization.” Gupta is using his identity as a resource with which to indicate what he can and cannot do. The successful resolution to the negotiation, and the fact that both can be seen to have acted in an organizationally defensible manner, confirms and maintains their identity as proficient organizational members.


This article has sought to show how various narrative and non-narrative resources are deployed in skilled activity to create highly valued social and moral orders from the ambiguity that exists in any one context, a movement from uncertainty to valued certainties. In particular, it has focused on the non-narrative and argumentative aspects of skill by describing the argumentative structure of a skill (i.e. the claim, warrant and grounds) and showing how this functions to legitimate the appropriateness of a narrative feature of a skill. Indeed, the argumentative structure of a skill plays a crucial part in whether actions are perceived as valued and legitimate, and thus take a crucial role in the creation of skilled activity.

It was additionally shown that, although the argumentative structure of an act is normally left unstated, it can be used proactively or reactively to justify an act. Furthermore, it is when there are differences in what is considered legitimate that the argumentative aspects of skilled activity become much more visible and important if valued social and moral orders are to be successfully constructed. Also demonstrated was the way in which narrative and non-narrative features of a skill are used reflexively to create the activity and context of negotiation and that this is a social accomplishment that occurs in a turn-by-turn sequence and in a way that is contextually situated and responsive to novel and contingent demands.

The dialogical approach has a number of implications for organizational practitioners. First, the dialogical approach draws organizational practitioners’ attention to the range of narrative and non-narrative features that are used in skilled activity. In particular, it shows how these are used to create consensually valued social and moral orders. However, as there may be a number of different understandings of what is valued in any one context, the dialogical approach stresses the need to pay continual and close attention to the intentions and perspectives of others. This implies that reflection should be focused both internally and externally, i.e. to reflect on one’s own understandings and on those of others.

It is interesting to note here that the analysis of arguments can help managers to uncover and better understand their own and others’ intentions, desires, perspetives and identity (see Gold, Holman and Thorpe, 1999, for a study of the use of story telling and argument analysis in management development). The dialogical approach also suggests that skilled activity may best be viewed as a joint social accomplishment that is achieved with others and not done to or for others. For practitioners, this means that they can think of themselves as coauthors of a context and not as someone who must apply the rules of skill to a relatively pre-given context (Shotter, 1993). Another implication of the dialogical approach is that, due to the context-dependent nature of skills, difficulties will often be encountered when generalizing a skill from one context to another. Indeed, what is seen in a skill in one context may not be so in others. This points to the basic fragility of skilled activity and to the fact that the line between artful and skilled activity is never fixed.

Perhaps the most important contribution of the dialogical approach to organizational practice is that it draws attention to the need to argue for the appropriateness and legitimacy of action. One can envisage practitioners creating “landscapes” of action within which a skill can appear valid, and developing persuasive arguments for the validity of those landscapes (Shotter, 1993). Such argumentative activity would also be critical when people’s understandings are different, when a skill is used in a new context, or when a person is confronted by a novel variation. In such circumstances it might be expected that more narrative features would be deployed and, crucially, that more argumentative effort would be needed if actions were to be viewed as legitimate and skilled activity were to be generated. It might also be expected that, in these situations, practitioners would pay much closer consideration to the joint authoring of mutually valued moral and social orders.

Finally, the dialogical approach implies that practitioners should be aware of the possible need to argue for the appropriateness and validity of a skill at times other than when the skill is used. This is because arguments that support the legitimacy of a skill may be made, for example in conversations or in books, that are “unconnected” to the actual use of the skill. Skilled activity may not, therefore, be fully understood without reference to these “unconnected” instances and this points to the need to understand the wider social and cultural context that supports the use of a skill.