Skip to content

Talk vs. Action: Conversation as Technology

Talking is often dismissed, especially in business, as something that is categorically inferior to doing. Why is that? And just what is the “doing” here referred to? The simple reason for this so-called “bias to action”, I think, is that within the economy of industrial production, talk is just grit in the slick grease of efficiency, which all too often is obssessed with “saving” time. Within such an economy, talking, which takes time, can only stall or slow decision and action.

When Elvis sang, “a little less conversation, a little more action, please,” he was hoping to hustle things up and get laid. Fair enough, so far as it goes. And I think this is just what business wants to do, hustle through or past the work of relationships and get straight on to the main event: the consummation, in this case, of the transaction. This, again, within a framework where efficiency and the minimization of costs are paramount, makes perfect sense. So the talking which is so much a part of establishing relationships, conversation, is marginalized by more efficient forms of communication, like orders, memos, and meetings (OK, we can all be forgiven our suspicions about the efficiency of that last one).

Casual conversations abound in social life. Serious, deep, conversation, however, is a kind of talking that happens far too seldom in general, and which happens only in isolation or in the margins in the context of business or work. What I mean by conversation, I should clarify, is quite particular. A conversation has multiple participants, engagement is voluntary, and the talking in a conversation is governed by interest rather than status or role. Conversation is a form of communication, then, that involves an exchange among the participants where the value of content is emergent and the interest in participation is intrinsic. In other words, we can never know at the start of a true conversation where it will end up and we give ourselves over to having conversations primarily because it feels good to have one.

What conversation gives up in terms of predictability and control, however, it makes up for in the value of its reflexive and looping design. What conversations can produce organically and emergently, I’d argue, are intelligent outcomes. The structure of conversations is neither that of truth trees or decision trees, nor of waterfalls or stage gates. Conversation is an irreducibly social form of communications and consequently shares the structure of the social, which is to say, it has the shape of a network. In other designs or patterns of communication, ones with essentially linear pathways, the intelligence is usually located at the front end, as for example in a command hierarchy. The regime of this method of communication is explicitly designed against revision, and so, by definition, orders cannot become more intelligent after they are given, and yet it is only after they are given that they start to be tested against the shifting dynamics of context and interactions. This matters least when the objectives of action are well understood: attach that fastener, take that hill, get me a coffee.

Conversations, on the other hand, have two structural elements that can augment or enhance the thinking of the participants: these are diversity and the “looping effect” of the network. These elements are not optimal by default in every conversation, but rather are structural characteristics that can be optimized. To put it another way, the features that regulate the quality of a conversation can be designed. The first element, diversity, derives from a principle of cybernetics: Ashby’s “law of requisite variety”. Simply put, the potential for intelligent outcomes to arise from a conversation are related to the participants collectively sharing a sufficient diversity to avoid simple reproduction (likemind) or too much disagreement (stalemate). The second element, what I call the looping effect of the network structure, is related to but not reducible to diversity. The concept of a feedback loop is well understood and permits us to explain how systems can learn and improve. Applying this framework to a network structure produces a looping pattern of dynamic potentials between the nodes of the network (these looping effects are a more granular phenomenon of network effects): in a human conversation we call these digressions, tangents, sidebars, and so on. But what can appear to be “time wasters” in the context of a more linear design for interaction, e.g. a meeting, here can give rise to the unexpected, which is often where the intelligence of the interaction leaks into the content.

Conversation is a powerful framework for prototyping scenarios or any other situation where wicked problems arise or are in play. This is because, if well designed and well guided, a conversation is a dynamically adaptive system of interaction, a system for enabling, capturing and looping distributed cognition and dynamic data. Edwin Hutchins, in his amazing study of oceanic navigation (Cognition in the Wild), at one point observes that a nautical chart is, in fact, an analog computer. At first, this seems odd and it requires a bit of time to loose the wrench it throws into our normal way of thinking about what a computer is or what a map is. On further reflection, however, we begin to understand that the object that we take for granted, the chart, contains many layers, implicates a number of technologies, metrics, rubrics, symbolic orders, and so on. All this, furthermore, still leaves out what is Hutchins’ main object of study, the social system that is required for such a device, a chart, to be anything other than an inert object. One of Hutchins’ key insights is that the web of interaction between the team of naval navigators is both a kind of emergent intelligence and that this intelligent is itself a component part (an underlying technological layer) of a system of navigation.

My point is that conversation is a technology that is far more sophisticated than we take it to be. It seems strange, just as in Hutchins’ description of a chart as a computer, to even call a conversation a technology at all. But I think it vital that we not only start to think about how the technologies of conversation operate, but to learn to design for them. I believe that conversation is a crucial platform in a group of emerging technologies that include: networks, intelligence, knowledge, peers, sharing, and, even…freedom.

Nota Bene: Too many people deserve credit for shaping my thinking on these. But particular thanks go to Robin Uchida, Dr. Paul Pangaro and the Overlap community.


  1. Adam Thody wrote:

    Michael, you raise some very good points. I might just add that, in business, the tendency to minimize conversation is because many of us simply aren’t “good” at it. Many of us are prone to going off on too many inconsequential tangents, which is usually a symptom of not being able to assess our audience. Some personality types are simply uncomfortable engaging in off-the-rails conversation with coworkers. Then there are those of us who like to take time, process inputs, then compose feedback at a later time.

    It’s because of these not-so-subtle differences in conversational style that most of us despise meetings. Since we usually don’t have the luxury of hand picking meeting participants, there’s a good chance that we’ll encounter a less compatible conversationalist, making the encounter more difficult to bear.

    Like you said, it’s a matter of context—there’s a time and a place. Given the proper facilitation and constraints, conversation is a very powerful, and oft under-utilized technology.

    Tuesday, July 21, 2009 at 4:52 am | Permalink
  2. Nicely done sir. This piece is a fantastic companion to the awareness I have gained in engaging in constructed conversation with you. There is much to the selection of participants and you nail that with the like mind and too much agreement dimensions. Ensuring some essential tension in conversation is critical to creating valuable content rather than the drivel most engage in that is labeled as conversation or more often labeled “talking”. This tension could be seen as an ‘arguement’ in some views but ultimately it can be anywhere from a simple fact check to calling into question the most fundamental values of a construct. We think in extremes and will almost surely see the purposeful introduction of tension as a negative but tension creates opportunity. Tension is the fundamental driver of valuable content.

    Anyway, am on iPhone so will cut this short.

    Recognizing conversation as technology is a more recent development and is aresult of knowing Michael. It has encouraged me to reach out to find those that can impart that essental tension rather than those who I am in complete agreement with. Those conversations have been fantastic for amplifying my opportunity to develop more meaningful content and opportunities.

    (apologies for errors or poor construction from commenting via iPhone)

    Tuesday, July 21, 2009 at 7:39 am | Permalink
  3. Jose Leal wrote:

    Michael, I think you’re right about conversation being necessary and powerful for a number of reasons. But, I would have never thought of it as a technology. To me conversation is about building on ideas and arriving at a mutual point that none of the parties started at. Then again maybe it is!

    Thinking of conversation as a technology brings with it a lot of different opportunities. That is, opportunities to utilise it as an integral part of problem solving rather then that of problem making.

    Building clear processes and functions around conversation would help to integrate it into scenarios that would greatly benefit from a well structured conversation. As with other technologies, I guess it comes down to having clear parameters, structure, expected inputs and outputs….hmmm!

    All in all, a very interesting post. Thanks!

    Tuesday, July 21, 2009 at 5:30 pm | Permalink
  4. Really interesting post, Michael.

    I agree that design of (and for) conversation is a crucial platform for innovation and change. Too little effort is spent on shaping conversations for the generation and exchange of meaning. I wonder: If we describe conversation as a technology, will that perspective shift inspire people to learn how to use it more effectively?

    One specific looping effect worth noting is that we not only respond to what others say, but to what we ourselves say. Hearing our thoughts aloud enables reflection and facilitates processing that is qualitatively different than what we reap from written exchanges. Listening to ourselves enables us to evolve our ideas, the language we use to articulate them, and the manner in which we convey them.

    I agree that the more we engage with both openness and intent, the more likely we are to experience value in conversations and, dare I say it, have fun. But, your phrase “we give ourselves over to having conversations primarily because it feels good to have one” doesn’t seem quite right. Deep conversations can be challenging, so I don’t think it necessarily follows that they feel good. Also, isn’t is possible to give oneself over to conversation because of the anticipated value rather than because of an emotional benefit?

    (I look forward to diving more deeply into these issues. Thanks for the inspiration.)

    Friday, July 24, 2009 at 3:49 pm | Permalink
  5. The notion of ‘Communication’ as a ‘Technology’ is a good one. I would push on the specific role it serves in a distributed system. The framework of distributed cognition draws attention to all sorts of material and immaterial factors that shape the way we think, act, and behave. Communication seems to be a vessel for uniting or sharing ideas between sharp minds, their experiences, and their worldly artifacts.

    But perhaps more concretely, casual conversations and serious, deep conversations play complimentary roles. When we meet up to chat informally (or when we share personal quips on Twitter), we gradually develop an appreciation for and a deeper sense of ‘knowing’ each other. This can directly lead to better serious conversations. Serious conversations, on the other hand, let people dive deep on some topic or issue — and by sharing their combined, diverse knowledge — they come up with creative solutions.

    Actually, casual versus serious conversations are similar to Rob Burt’s notions of brokerage versus closure. Burt studies how ’social capital’ gets formed in workplaces and found that managers who engaged in lots of ‘brokering’ — by engaging widely with people or groups from different backgrounds — return to their primary workgroup with more/better knowledge. These managers earn more money, get more promotions, roll out more products, etc. Closure, however, compliments this ’spreading wide’ behavior by bringing ideas back to a central place and pushing deeply on them, much like we do in serious, technical conversations. In the end, both are necessary for productivity. In the end, both types of conversations are necessary for productivity, too.

    Thanks for your great post, Michael :)

    Thursday, July 30, 2009 at 11:03 am | Permalink
  6. I just came across an article from the Jul/Aug Interactions magazine on: “What is conversation, and how can we design for it.” Thought it might be relevant to share:

    Saturday, August 1, 2009 at 12:31 pm | Permalink

Post a Comment

Your email is never published nor shared. Required fields are marked *