Speech Acts and Promises--Conversations for Business

Austin, Searle, and Flores meet Promise Theory

Tue 4 Aug 18:03:44 CEST 2020

Years ago, during the genesis of Promise Theory, one of the many books I read and set aside was John Searle's book on Speech Acts. As a physicist by nature, and a Computer Scientist by trade, I'm wary of a lot of philosophy, which often feels at odds with the rigour of more formal theory, but I always commit to doing my homework, and there are more and less rigourous philosophers. Ideas are ideas, and one should always show a willingness to be receptive to them, without necessarily swallowing everything from the firehose.

Earlier this year, I was encouraged to revisit the philosophy of speech acts by friends and colleagues from the Open Leadership Network, when I learned of the subject from an unexpected angle: it turned out that John Searle's student Fernando Flores had written on this subject too, and applied it to study the language of business dealings---with its own following. Flores' essays, privately held from the 1980s onwards, were collected in 2012 and published in a volume called Conversations for Action: Instilling a Culture of Commitment in Working Relationships by his daughter Maria Flores Letelier---sparking a new following in the business world (always hungry for ideas about leadership).

Language for action

The theory of Speech Acts began as an investigation into the meaning of certain expressions in language, in post war Oxford University. That might sound like the typically fussy domain of philosophers, quibbling over words, but it turns out to be importance both in business and in technology. Speech Acts specifically take on the issue of how certain phrases are not merely descriptions of actions, but are in fact actions in their own right--how conversation is itself an activity that drives processes, whether in business for getting things done, or elsewhere as a way of shared reasoning, negotiating, and directing activity. Utterances (or speech acts) can be classified by their active function, not just by their grammatical status.

Making promises is one such example. Promises are the basis of cooperation and contract. There are many others as discussed by the progenitor of Speech Acts: John Langshaw Austin, in his book How to do Things with Words (1955 revised 1962), and followed up by the more famous John Searle in his book Speech Acts: An Essay in the Philosophy of Language (1969), then later by Fernando Flores in a number of essays during the 1980s.

One of the reasons I set aside many prior works on the philosophy of promises, and contracts, when developing modern Promise Theory, was due to the pervasive assumption that intent had to be associated with obligation, rather than with the autonomous decisions and actions of individuals. (Consider that sentence! Intent MUST be associated with obligation -- it's obligatory!) This curious obsession with thou shalt (must, should, obliged, etc) rather than we intend to try haunts both philosophy and daily life, leading to much confusion. It probably arises from the Western biblical tradition. The language of subtle threats, of law by decree, or deontism, is by far the most common explanation of intentional language in philosophy and logic. But, when we look closer, there is always a deeper explanation in terms of voluntary cooperation actually at work behind the facade. Perhaps this is evidence of our long history of subjugation by leaders, religions, and political forces over the millennia. The following quote illustrates how an unspoken undercurrent of obligation is implied in nearly everything written before Promise Theory.

The process of communication should be designed to bring with it a major awareness about the occurrence of commitments. Every member's knowledge about his participation in the network of commitment must be reinforced and developed.

-F. Flores, Management and Communication in the Office of the Future, 1980

The passage from Fernando Flores, student of John Searle, strives to express an autonomy, but equally feels trapped by the latent assumption of a virtual obligation that was so entrenched at the time. It's hard to escape endemic convention. Various attempts to smooth over the language, to make the assumption of obligation less aggressive are explored by Searle and Flores---yet the basic implication remains, and the discomfort of conflating someone else's intent with voluntary action is never quite resolved. But now we know better.

Promises are distinct from obligations

What Promise Theory really dared to point out, in a satisfactory way--by following the simple pattern of interaction theory from physics, is that promises are (local) expressions of intent from within the agents that makes them, whereas impositions (which include the attempt to induce obligations) are directives from without---exterior agents who would force or induce influence in others. In Promise Theory, the rational starting point is that: no agent (human or otherwise) can make a promise about anything or anyone other than themselves. One could dispense with the legacy of deontism altogether. Obligation is the default way we phrase laws, from Moses until today, but even the most ardent lawmakers know that imposing a law, which people don't basically promise to abide by in the first place, is futile. The widespread commonplace understanding of law as involuntary is entirely misunderstood.

Similarly, as physics saw a subtle shift away from the concept of force and control to message and influence between autonomous (local) entities during the 20th century, so a similar shift is nascent in the philosophies of several authors, but only Promise Theory seems to grasp the distinction fully, especially the reason why that is important.

A distinction that Speech Acts doesn't make very well (perhaps due to the muddle about obligations) concerns the difference between an imposition (attempt to induce cooperation) and a promise (voluntary intent). When expressed up front, "expectations" and "requirements" (so commonly thrown around in business-speak) may be seen to fall under the category of impositions, because they may challenge others to comply, with a latent threat hidden just under the breath. Promise Theory predicts simply that cultures based on this kind of imposition are somewhat risky and toxic: "expectations" and "requirements" that are imposed (and not refuted) will eventually become perceived by an imposer as promises that were actually made (taking liberties). An example of this can be found in the Boeing 737-Max debacle. And there is an entire corollary about blame and responsibility that follows on from those simple notions. Promise Theory has been highly effective at elucidating those dark issues.

In "Scrum", the word "commitment" was used to describe what the team comes up with at Sprint planning time. But, over time, that ended up getting weaponized by management, and used to punish teams who did not "meet" their "commitment. From Best Effort to Mandatory Obligation.

--Daniel Mezick

Getting it right ... mostly

Searle and Flores spend a lot of time categorizing words into different kinds. Does a phrase imply an obligation? Is it a statement of description or intent about the future? How does the listener perceive what the speaker or writer intends? Many of the distinctions they made I find unimportant and unconvincing. For instance, in their view, promises could only be about future actions. This does not seem to be true, as I wrote in the Promise Theory book by counterexample: `I promise that I fed your cat yesterday' is certainly a promise, yet about the past. The point is not that the action is in the future, but rather than the outcome of the promise is yet to be assessed---assessment lies in the future. It's about a lack of knowledge, not about a timeline of who did what. The role of time somehow got muddled in these texts---Promise Theory simplifies those distinctions.

Conversely, Flores carefully underlined the importance of trust in passing on intent between agents of language, which several others missed, and properly separated assessments from statements of intent. There is a lot of overlap between Speech Acts and Promise Theory, but there is also a key distinction.

Whereas Speech Acts (in typical philosophical style) attempt to discriminate every role and nuance, and attach new names to each subtle distinction, Promise Theory (in typical physics style) attempts to condense the major albeit approximate distinctions that underpin roles during interaction---to find the 20% of structure that explains 80% of the phenomena, rather than the other way around.

When do we need our humanity?

Moving away from a humanist view of activity was the key to industrialization, and although we are now becoming post-industrial, an important lesson was learned about the tradeoff between individual freedoms and the consistent discipline needed for cooperation as groups, troupes, companies, and societies. Sometimes we need to set aside our humanity in order to work as a unit. An orchestra is a good example. Brilliant players voluntarily give up their individuality to come together as a single machine. Without that voluntary promise to subordinate themselves to a larger goal, orchestral performance would be impossible. Today in our `Age Of Plenty And Order' we sometimes indulge in an exaggerated recognition of the virtues of individuality and liberty, de-emphasizing discipline and machine-like efficiency for an individual concept of human well-being. That's how social groups fall apart.

Focusing too much on human agency prevents us from seeing a simple truth: we often act (involuntarily) as a conduit for others' intentions! We are not always aware of our own intentions in everything we do.

As the pendulum has swung from wartime collaborative discipline to peacetime individuality and creativity, the concept of voluntary subordination for a greater good has been swept aside with modern indulgence. Today, we can replace humans acting as machines with actual machinery in many cases---but that's not always a good investment for keeping the wheels turning. An orchestra is the go-to example of a collaboration where that would be both impractical and undesirable--it relies as much on human nuance as it does on mechanical discipline. In any cooperation, there has to be a balance between these two strategies for working. Without discipline, action cannot scale; without individuality, innovation and wellbeing cannot thrive.

Flores spend a lot of time on human moods as contexts for linguistic nuance, and explaining how language could lead to misunderstandings. He described effective protocols for bringing up difficult issues, and bringing a sense of completion to conversations. I think he had many good points that have since informed other writers. Promise Theory doesn't exclude these issues, but also doesn't try to go there. It stays more on the impartial mechanical side of `necessary and sufficient'. If we get those basics right, we can develop nuance later. That gives us a simple bottom line, underlining the fragility and potential failure modes of a collaboration. Better to keep core principles independent and `inhuman', than to muddy the waters with emotional aspects, because when offered the chance to use emotion over reason, most of us go with our guts. Emotions may, of course, lead to signalling of promises through body language and other behavioural patterns. However, it's not clear whether a formal language, like Promise Theory, would be the best way to handle those issues in a business environment. In A.I. it could be a quite different story---but for another time.

Less nuance, more pattern matching

Is language just a matter of taste or are there functional distinctions? That's what Speech Acts asked. How much and how little can we get away with? As intentional beings, precision in meaning is important to us. In collaborative activity, where critical outcomes depend on effective communication, it's not surprising that there is a lot to be gained from optimizing communication. Austin introduced specialized terms like locutionary acts, illocutionary acts, and perlocutionary acts. The story was rewritten in a practical vein, to identify:

  1. Requests or offers.
  2. Promises or Acceptances.
  3. Declarations of completion.
  4. Declarations of satisfaction.

Clearly one could keep going. Counter-offers to offers can be added to protocols for negotiating, and commissives or commitments rake up the old obsession with obligations--suggesting a building of trust. But as language distinctions become increasingly fussy (especially in seeking rule-based grammar), we lose the virtues of approximation for bundling meaning and expressing at scale. Cooperation and harmony may end up being exploded on this alter of precision. The scaling meaning is what Promise Theory tries to achieve. There is no need to distinguish requests and offers from promises in most cases. They are both just declarations of intent, with inconsequential differences from a perspective of necessity. Promise Theory simplifies the story:

  1. Promises to offer (+), promises that accept (-) -- useless without one another.
  2. Impositions that try to induce acceptance or service.
  3. Assessments about promises (kept, not kept, etc)
  4. Conditional promises: I promise X if you promise Y.

From this smaller set of distinctions, remarkably, there is in fact enough structure to understand the core patterns of how organizations work as machinery, without rejecting human nuance altogether:

  • Agreement
  • Common knowledge
  • Rights, permissions, and privileges
  • Service interactions
  • Authority and authorization
  • Invitation versus imposition
  • etc

Engineering Cooperation

Promise Theory concerns itself with what follows on from those signals assuming that they are understood. It comprises a number of principles by which we can formulate The Engineering of Cooperation---not just between people, but in all interactions, between animate and inanimate parts. That includes processes for negotiation and agreement, not just command and control-- and when to use those patterns. What may surprise many (thanks to the way language is taught in schools, and is often worshipped as the summit of human agency) is that a lot of meaning lies in the interactions between agents, not in the precise representations of language. As semaphores and code books prove, meaning can be conveyed without fussing over the roles and meanings of individual words. Indeed, it can be argued that meanings don't lie in words at all, but only in complete utterances (usually sentences). Single symbols can be imbued with complex meanings. How those meanings come to be agreed are an entirely different issue.

Promise Theory doesn't concern itself with the minutiae of meaning, as Austin, Searle, and Flores did (not because that isn't important, but because it isn't always the simplest way forward), but rather it probes the `mechanics' of what's necessary and sufficient about signalling, more like physics. Language is just one way of representing intent---and when you strip away that assumption, there remains a skeleton whose underlying semantics can still shed light on interactions---human and machine. By focusing on promises, we focus on actionable roles that we do and can control, autonomously

Flores' operationalization of language has inspired many in the business world, the realm of cooperation and leadership, but it's not easily formalized. Promise Theory is a simpler story that can be formalized, and to good effect. It predicts simple patterns for successful interaction: the simple basis of authority, service, permission, privilege, the meaning of invitation, and much more.

MB, Tue 4 Aug 18:03:44 CEST 2020

Acknowledgement: I'm grateful to Daniel Mezick for comments and the reference to the Scrum Guide.

Training Course in Basic Promise Theory

Learning how to transform thinking from obligation to promise.