Protocol formalism and antiformalism
16 Oct 2026 - 17 Oct 2026
- In my efforts to work with the protocol community, I found myself drawn to the Formal Protocol Theory SIG (I do have a mathematics degree after all), but also found myself not quite in sync. This is nothing to do with them; it's due to a long-standing set of intellectual biases of mine.
- I learned this from David Chapman and Phil Agre, who launched a crusade against some of the fundamental assumptions of GOFAI.
- Hm, need some kind of a diagram GOFAI, neural nets/ML, logic (McCarthy) Minsky, Situated Action, Cybernetics maybe.
- Well the upshot is, I don't trust formalisms. They necessarily abstract away from reality, and while you have to do that, you should do it in a more flexible way than math usually supports.
- Of course the whole thrust of the protocol movement is to unite formalisms and reality – this is kind of what New Nature means. So I may just not be down with the program.
- When I started out I was more interested in natural, informal protocols, such as those found in face-to-face human interaction. The interesting fact about this kind of interaction is its lack of foundational formalism. Ethnomethodology is based on this lack, and focuses on how things that are sort of like formal knowledge are crafted by the participants in real-time.
- This must be an explored topic, give me a literature review
- Ask Claude
- This is a rich and genuinely well-explored area. Here is a structured literature review covering the main threads.
Literature Review: Formalism, Antiformalism, and the Informal Protocols of Face-to-Face Interaction
1. The Core Antiformalist Claim: Order Without Foundations
- The intellectual tradition you're drawing on begins with a fundamental challenge to the idea that social order is generated by formal rules or representations.
- The point of ethnomethodology, roughly, is that the formal structures of protocols are not constitutive but are the product of people's improvised social activity.
- This inverts the usual picture: rather than rules producing behavior, behavior produces the appearance of rules.
- The canonical source is Harold Garfinkel, Studies in Ethnomethodology (1967).
- According to Garfinkel, ethnomethodology assumes that the meaning of a social phenomenon is equivalent to the methodical procedures through which participants build and maintain its sense. It aims at revealing the primary features of social actions by discovering their foundation in situated practices at the face-to-face level. Social order is not given top-down, but something that participants work to achieve.
- Garfinkel's famous breaching experiments were the empirical method for revealing this.
- Garfinkel's central argument was that social order is not imposed from above through explicit rules and sanctions, but remains hidden and actively produced through members' unspoken, taken-for-granted background expectancies — what he called the "natural facts of life."
- Garfinkel developed methods for deliberately breaching social order to study how people work to maintain what they expect of it — pedagogical tools used to explore "the persistence and continuity of the features of concerted actions" of everyday life.
- The diagnostic power is negative:
- social order is an achievement; it holds because people produce it together, moment by moment, by honouring expectations that can't be comprehensively documented.
2. Conversation Analysis: The Discovery of Informal Structure
- The most productive empirical offshoot of ethnomethodology is **Conversation Analysis (CA)**, which found real, discoverable structure in informal interaction — but structure that is locally produced, not pre-specified.
- Conversation analysis, the most visible form of ethnomethodological research, is concerned with describing the methods by which members of a culture engage in social interaction. A key goal of CA is to examine social interaction to reveal organized practices or patterns of action, under the fundamental assumption that interaction is structurally organized.
- The founding paper is **Sacks, Schegloff & Jefferson*, "A Simplest Systematics for the Organization of Turn-Taking for Conversation," Language* (1974).
- The authors show that in the course of a conversation, co-conversationalists use a number of techniques to organize turn-taking: transfer of speakership by the current speaker allocating the next speaker; self-selection by a participant; or continuation by the current speaker. The result is an orderly conversation based on the principle "one speaker at a time."
- The crucial point is that
- in every conversation there is a turn-taking system in operation which provides for a flexible adaptation of the conversation's structure to any possible topics and any possible speakers' identities.
- This is the antiformalist's ideal: structure that is flexible, locally adapted, and participant-produced — not pre-written.
- Rather than asking what an interaction meant to the people in it, ethnomethodology asks how they made it observably, publicly orderly.
- A key theoretical tension within EMCA (ethnomethodology + CA) is itself over formalism.
- Critiques regarding unwarranted formalism and "scientized pretensions" have been directed against leading conversation analysts including Schegloff and Sacks himself
- — that is, even the relatively minimal CA formalisms (adjacency pairs, turn-construction units) attract antiformalist critique from within the tradition.
3. Goffman: The Interaction Order as Informal Protocol
- Running parallel to the Garfinkel/Sacks lineage is Erving Goffman, whose work maps the implicit normative infrastructure of face-to-face life without reducing it to explicit rules.
- Goffman's Interaction Ritual (1967) is a study of face-to-face interaction in natural settings — that class of events which occurs during co-presence and by virtue of co-presence. The ultimate behavioral materials are the glances, gestures, positionings, and verbal statements that people continuously feed into situations, whether intended or not.
- Goffman addresses social order at the level of the interaction order, in a quite different manner than structural sociology. Rules and procedures are more tacitly agreed upon, coming from society but individually interpreted, and maintained through tacit procedures in social encounters.
- Goffman's key insight, relevant to your protocol interest, is that
- face arises from the fit between expectations and behaviour: face-work is rule-following, the "ritual" to which the title of the book refers
- — but rule-following in a sense where the "rules" are reconstructed in real time, not consulted in advance.
4. Suchman: The Definitive Bridge to AI and Protocol Formalism
- The most directly relevant figure for your intersection of ethnomethodology and AI/protocol formalism is Lucy Suchman, Plans and Situated Actions: The Problem of Human-Machine Communication (Cambridge University Press, 1987; revised 2007 as Human-Machine Reconfigurations).
- This book offers a provocative critique of the dominant assumptions regarding human action and communication which underlie recent research in machine intelligence. Suchman argues that the planning model of interaction favoured by the majority of AI researchers does not take sufficient account of the situatedness of most human social behaviour.
- Suchman argues that real situated human action cannot be solely based on plans or internal representations.
- Instead, real-world actions often deviate from pre-designed plans and must be adapted to specific, local, contingent determinants. From this perspective, "plans are the resources for situated actions" rather than prescriptive instructions — orienting actors' behaviors but not fully determining their conduct.
- Suchman's early research was heavily influenced by ethnomethodology. Her book provided intellectual foundations for the field of human-computer interaction.
- It is the fulcrum text for your concern: the place where the ethnomethodological critique of formalism meets AI and protocol design head-on.
5. Agre and Chapman: The GOFAI Critique from Inside AI
- Your mention of Phil Agre and David Chapman is well placed. They launched their critique from inside AI, not from social science.
- Agre and Chapman initiated the "situated action insurgency in AI," a movement that questioned the foundations of the stale theories of mind that were implicit in the computational approach.
- Technically,
- Agre and Chapman criticized purely plan-based accounts and showed how routine activity depends on situated forms of interaction and deictic representation.
- Their system Pengi demonstrated that
- the world of action is made up not of present-at-hand objects with properties, but of possibilities for action that trigger appropriate responses from the agent — what Agre called "deictic representations."
- The philosophical backdrop here is Heideggerian:
- Agre proposed that "AI's whole mentalist foundation is mistaken, and the organizing metaphors of the field should begin with routine interaction with a familiar world, not problem solving inside one's mind."
- Alright but that was mostly stuff I'm already familiar with, I wanted to know if there's related stuff from (eg) philosophy of mathematics.
- Ask Claude
- This is a rich vein. Let me pull together the key threads from philosophy of mathematics that connect to your antiformalist intuitions.
Antiformalism in Philosophy of Mathematics: A Literature Review
- Your position has substantial, well-developed parallels in the philosophy of mathematics, largely under the banner of **Philosophy of Mathematical Practice (PMP) and the "maverick" tradition**. Here's the landscape:
1. The Maverick Tradition (the closest analogue to your stance)
- Aspray and Kitcher identified two traditions in philosophy of mathematics: the mainstream, starting with Frege and centered on foundations, and a maverick tradition that began with Lakatos.
- The mavericks are characterized by three key features: (a) anti-foundationalism — mathematics is a fallible activity with no certain foundation; (b) anti-logicism — mathematical logic cannot adequately analyze mathematics and its development; and (c) attention to practice — only detailed analysis of actual mathematical practice can produce a philosophy of mathematics worth its name.
- The key figures:
- Imre Lakatos (Proofs and Refutations, 1976):
- Lakatos argued that "informal, quasi-empirical mathematics does not grow through a monotonous increase of the number of indubitably established theorems, but through the incessant improvement of guesses by speculation and criticism, by the logic of proof and refutation."
- He criticized Hilbert's formalist school, finding that formalism "denies the status of mathematics to most of what has been commonly understood to be mathematics, and can say nothing about its growth."
- This is almost directly parallel to your point about formalisms abstracting away from reality.
- Thomas Tymoczko:
- His anthology New Directions in the Philosophy of Mathematics gathered contributions explicitly aligned with maverick features, and its preface criticized dominant foundationalist perspectives as unable to "articulate the actual experience of mathematicians."
- Reuben Hersh (What Is Mathematics, Really?, 1997):
- Most philosophers of mathematics treat it as isolated, timeless, ahistorical, and inhuman; Hersh argues the contrary — that mathematics must be understood as a human activity, a social phenomenon, part of human culture, historically evolved, and intelligible only in a social context.
- His humanistic position is, in essence, that "mathematics is what mathematicians do."
- Hersh self-identifies as a maverick:
- he divides philosophers into "Mainstream" and "Humanists and Mavericks" — the latter see mathematics as a human activity, the former as inhuman or superhuman.
2. Wittgenstein on Rules and Formalism
- This is probably the most direct philosophical connection to your ethnomethodological intuitions.
- Wittgenstein held as a constant that "mathematics cannot be separated from a language and a human practice."
- The key move:
- The meaning of a rule is not an abstract object detached from practice, but is expressed in the way the rule is actually employed within a language-game.
- Rule-following reveals a deep connection between meaning, action, and social practice. A rule does not determine its applications through a hidden private mechanism; rather, its normativity arises from its role in a public practice governed by training, use, and communal standards of correctness.
- This is essentially an argument that formal rules are always already grounded in informal practice — which is exactly what ethnomethodology says about social interaction. Kripke's reading of Wittgenstein sharpens the point:
- following a rule is a practice which is essentially communal — "what it is for a person to be following a rule, even individually, cannot ultimately be explained without reference to some community."
- The Wittgensteinian argument also has a direct limit implication for formalism:
- questions about logical formatting ("why just these rules?") cannot be meaningfully formulated within the system, since it is precisely the application of logical rules which enables the formulation of any question; and the celebrated infinite regress argument on rule-following seems to undermine any explanation of deduction as relying on a logical argument.
3. Philip Kitcher's Naturalism
- Kitcher proposed that a mathematical practice has five components: a language, a set of accepted statements, a set of important unsolved questions, a set of reasonings used to justify accepted statements, and a set of mathematical views about how mathematics should be done.
- He claims we can regard the history of mathematics as a sequence of rational changes in mathematical practices, connected through a chain of "interpractice transitions."
- Kitcher holds that the esoteric abstractions of modern mathematics can be viewed as an evolved product of changing theories developed from roots in "unproblematic entities" — primitive, empirically grounded practices of human beings performing everyday operations on their environment.
- This is a grounding move similar to what you want: formal structures bottoming out in informal praxis.
4. David Bloor and the Strong Programme (Sociology of Knowledge)
- From the sociology side:
- Bloor's radical claim in Knowledge and Social Imagery was that even "hard sciences" such as physics and mathematics are dependent on social factors such as conventions, interests, traditions, and prestige, as much as on observable phenomena or abstract logical necessity.
- The Strong Programme was also explicitly inspired by Wittgenstein's argument about rules — that to apply a rule, a judgment of similarity or difference is needed.
- This is directly comparable to ethnomethodology's claim about informal, in-situ construction of formal-seeming knowledge.
5. Husserl's "Origin of Geometry" and Sedimentation
- Perhaps the most sophisticated treatment of how formalisms emerge from and obscure informal practice is Husserl's late essay.
- Husserl argues that even the invariant meaning found in theoretical disciplines like geometry has a historical becoming — through gradual abstraction and stabilization. This process allows for the reception of established meaning, which simultaneously entails the forgetfulness of the experiential basis and intuitive knowledge that made ideality possible in the first place. Husserl calls this Janus-faced dialectical process between discovery and forgetfulness sedimentation.
- Husserl uses this geological term to describe geometry as a linguistically mediated genesis of conceptual knowledge — human practice that can only be transmitted to other generations by expressing it linguistically and recording it in writing.
- Derrida's introduction to the text then uses sedimentation as a launching pad for grammatology.
- This is strikingly close to what ethnomethodology says: formalism is a sedimented accomplishment of prior informal activity, and the formalism tends to erase the traces of that activity.
Summary Map
- | Figure | Key Claim | |
|
| | Lakatos | Math grows by proofs & refutations, not axiomatic deduction | | Wittgenstein | Rules get their meaning from practice, not from abstract structure | | Hersh | Mathematics is what mathematicians do; it's social and historical | | Kitcher | Mathematical practice evolves through interpractice transitions from informal roots | | Bloor | Even mathematics is socially conditioned all the way down | | Husserl | Formal geometry is sedimented from pre-formal practice, which it forgets |
- The Husserl piece especially seems like a direct bridge to your ethnomethodological concerns — it's basically a phenomenological version of the claim that formal protocols are locally and historically produced, and the production