The Ladder, the Lens and Self-Talk
People who disagree in a governance setting are usually standing on different rungs of different ladders. This concept extends Chris Argyris’s ladder of inference with three layers that shape how anyone climbs it: the substrate, which decides which data exists before anyone looks; the lens, which a person inherits by belonging and which decides what they notice and what it means; and self-talk, the private narrative in which conclusions harden into beliefs. Read together with TIME, the three layers explain why disagreement in the Emergent form is hard to resolve, and they indicate which remedy fits which kind of disagreement.
The ladder
Argyris developed the ladder in his work on organizational learning, and The Fifth Discipline Fieldbook (Senge et al. 1994) made it widely known. It describes the steps between an event and a response:
- Observable data: what a camera would record.
- Selected data: the fraction a person notices.
- Added meaning, drawn from culture and personal history.
- Assumptions built on that meaning.
- Conclusions.
- Beliefs about how the world works.
- Actions taken on those beliefs.
The climb is fast enough to feel like a single act of perception. Its most important feature is the reflexive loop: beliefs at the top decide which data gets selected at the bottom next time, so a belief can seal itself against correction. Argyris’s remedies work on reasoning. Make each rung visible, walk others down their ladders, and test assumptions against directly observable data.
Three layers
| Layer | Scale | Rungs | What it does |
|---|---|---|---|
| Substrate | Structural; shared by everyone using the same instruments | 1 to 2 | Decides which data exists and reaches anyone at all |
| Lens | Collective; inherited by belonging | 2 to 3 | Decides what a person notices in the available data and what it means |
| Self-talk | Individual; private | 4 to 6 | Builds assumptions, conclusions and beliefs, and tells the result back as plain seeing |
Each layer shapes the one inside it. The substrate supplies the data a lens works on, and the lens supplies the meanings self-talk works on.
The lens
A person acquires a lens by belonging: to a coordination form, a community, a profession, a language or a shared history. People who share a lens rarely see it, and they tend to finish each other’s sentences. Everyone carries several and switches between them by setting. A commercial farmer may read a water allocation through a Market lens at the farmers’ association and through a Tribal lens at church on Sunday. The four TIME forms are one family of lenses; faith, class and discipline are others.
Read as lenses, the four forms supply four different filters for selection and meaning:
- T form: who is kin, who belongs, what honor demands.
- I form: who holds authority, what the rules require, what counts as legitimate.
- M form: what it costs, what it earns, who will pay.
- E form: who knows what, where information flows, what the network can do together.
Take one observable event: a farmer clears a wetland to plant. Kin see a family providing for itself. The municipality sees a zoning question. The market sees an input cost and a yield. An Emergent reading sees a hydrological node that dozens of other actors in the catchment depend on. Each reading holds together from the inside. Conflict in a valley often turns out to be lenses talking past each other at rung three while their holders argue about rung five.
Self-talk
Self-talk is the individual’s running account of what is happening, who is responsible and what comes next. It sits on the upper rungs and carries the reflexive loop: self-talk remembers the conclusion, and the conclusion tells the lens where to look next time. It also absorbs the substrate. “I worked for what I have” is an M-form conclusion experienced as a personal truth.
Cognitive biases are self-talk’s working tools. Four do most damage when perspectives meet:
- Naive realism (Ross and Ward 1996): I see the world as it is, so those who disagree must be misinformed or acting in bad faith. It turns a lens difference into a judgment of character.
- Reactive devaluation (Ross and Stillinger 1991): a proposal loses value because of who tabled it.
- The ultimate attribution error (Pettigrew 1979): my group’s failures come from circumstance; theirs come from their nature.
- Confirmation bias: the reflexive loop itself.
Two differences between lens and self-talk matter for governance. A lens belongs to a group, so naming one in public costs nobody face, and naming it tends to legitimize it. Self-talk belongs to the individual, and exposing it carries risk that falls hardest on whoever holds least power in the room. Lens disagreements concern meaning; self-talk disagreements concern other people’s motives.
The substrate climbs the ladder first
The Substrate Hypothesis adds the layer Argyris’s tradition leaves out. Argyris located the problem in participant reasoning. The Substrate Hypothesis predicts that better reasoning habits fade, because the coordination substrate performs the lower rungs before anyone starts thinking.
Money is the clearest case, because price selects data. Anything with a price reaches rung two automatically for everyone who uses the currency. Anything without one, such as moisture held in the soil or a working riparian strip, drops out before interpretation begins. The M-form substrate has already done the selecting and supplied the meaning (value as exchange value), and it delivers a ready-made conclusion: more throughput is better.
This gives the Cheapest Available Behavior Thesis a cognitive side. The cheapest behavior goes with the cheapest inference, which is the one the substrate has already drawn. A workshop can expose that preselection for an afternoon; the next morning the rates bill and the bank balance do the selecting again.
The Coercion Continuum reads as the reflexive loop at civilizational scale. Each phase turns a conclusion (“labor must be compelled,” later “people must earn money to live”) into a belief that no longer registers as a belief. Once naturalized, it selects the evidence that confirms it, and alternatives look naive because the data that would support them never reaches rung two.
The (+T+I+M)^+N test in ladder terms
A +N effort can run excellent dialogue practice and still reason entirely over data the M-form substrate chose. That produces triform coordination with a +N overlay: collaborative conversation about facts that price selected. In ladder terms, the diagnostic question is whether the group, with the monetary substrate removed, still has a shared source of rung-one data that it trusts and acts on.
This is why the BioConomy puts weight on measurement infrastructure:
- Performance-based water bonds pay on measured hydrological outcomes. The payment trigger sits at rung one, which leaves little room for inference to drift.
- The BioRegion atlas and BioScore build a data layer selected on ecological criteria, changing what reaches rung two for everyone who uses them.
- Stigmergy coordinates through traces in a shared environment. Agents act on what they can observe, with almost no ladder in between.
- A BioHub coordination surface publishes offers, seeks and shared instruments in a common format, so partners start their ladders from the same data.
Protective illiquidity belongs here too. A bioregional unit that cannot be converted into the external unit of account resists being re-read through the M-form ladder. Once a commons asset becomes fungible, the price filter takes over rung two again.
Why the Emergent form needs this
Each older TIME form carries a built-in tie-breaker. In the T form the elder settles it; in the I form, the office-holder; in the M form, price, and nobody has to agree on meaning at all. The E form has none. When perspectives meet as equals, disagreement can only be resolved by working through the layers, so lens and self-talk work becomes part of how the E form governs itself.
The Substrate Hypothesis adds a caution. An E-form body usually runs on the same substrate as everything around it (public budgets, grants in the national currency, land under inherited tenure), and that substrate reinstalls its conclusions between meetings. Lens and self-talk work therefore has to be repeated. It keeps people working together long enough to build the shared instruments that will make much of it unnecessary: scaffolding for the period before the substrate changes.
Locating a disagreement
When a disagreement stalls, locate its layer first, because each layer takes a different remedy.
| Layer | Symptom | What resolves it |
|---|---|---|
| Substrate | People argue from different facts; one side sees data the other never does | Shared rung-one data: a site visit, a common map, a measured outcome, a register both sides read |
| Lens | Same facts, different meanings; each reading holds together from the inside | Name the lenses aloud; let each reading stand; look for options that satisfy more than one lens |
| Self-talk | Same meaning, still no agreement; the argument is about motives and trust | Time, relationship, private reflection, and separating proposals from proposers |
Stalled governance usually mistakes a self-talk disagreement for a lens disagreement, or a lens disagreement for a data disagreement, and applies the wrong remedy. More data does nothing for reactive devaluation.
Practice
- Involve people from the start. A finished proposal arrives at rung five, with the data already selected and the meaning already assigned through someone else’s lens. Involving affected people at rungs one to three is also a structural defense against reactive devaluation, because a proposal co-authored across lenses has no single other side to discount.
- Share data before debate. Circulate the rung-one material before each substantive item.
- Hold a lens round. Before debate, each party says what it notices in an item and what it means from where they stand, with no rebuttals during the round.
- Keep self-talk private. Argyris’s left-hand column exercise (write down what you thought but did not say) works alone or in pairs. Asking for it in plenary puts the risk on the least powerful.
- Acknowledge before deciding. Each speaker names the strongest point they heard from another lens. Sherman, Nelson and Ross (2003) found this reduces false polarization, while restating one’s own position does not.
- Separate proposals from proposers. Table proposals through a champion from another sector, or rate them before revealing their source.
- Teach the model. Learning about naive realism increased openness to the adversary’s narrative among the least open participants in one conflict study (Nasie et al. 2014).
- Record lenses. Note in the minutes which lenses were present on each item, next to what was decided.
What the concept does not claim
- It offers no substitute for substrate change. Lens and self-talk work is maintenance for the period before the substrate changes.
- It types no one. Forms and lenses describe the settings people speak from, and every person holds several. The research brief A TIME Profile for Multi-Perspective Governance sets out the evidence that type labels can themselves become lenses, and recommends domain-sensitive profiles and unscored tools.
- It treats no lens as more developed than another. TIME forms coexist and accumulate; ranking them would import a hierarchy of persons.
Open questions
- Are lens and self-talk layers of one model, or two diagnostic tools? The layered reading is tidy, but self-talk also decides which lens a person reaches for in a given setting, so the boundary leaks.
- Should a governance body use the vocabulary aloud? “Lens” works in plain speech. Named biases, spoken across a divided room, can sound like accusations and may belong in the facilitator’s toolkit.
- TIME has no form for the position grid-group cultural theory calls fatalist: people bound by rules they had no part in making. Where such people are present, how should the model register their lens?
Attribution
The ladder of inference is Argyris’s. Lens and self-talk entered the model in a Zoom dialogue between Michael Haupt and Michael Keller (Ecology of Design in Human Systems) on 25 September 2026. The substrate, lens and self-talk layers, their placement on the ladder, the disagreement locator and the application to TIME are Michael Haupt’s construction (working hypothesis). The four TIME forms build on David Ronfeldt’s TIMN framework; the reading of the fourth form as Emergent is the corpus’s own (see E form (Emergent)). The bias research is cited for its own findings.
See also
- Ladder of Inference, Lens and Self-talk. The glossary entries.
- A TIME Profile for Multi-Perspective Governance. Whether lens orientation can be measured, and how to make lenses visible without scoring people.
- Substrate Hypothesis. The parent thesis for the substrate layer.
- Cheapest Available Behavior. The behavioral thesis this concept gives a cognitive side.
- Consensus Is Not the Bottleneck. The paradigm-level argument about consensus; this page concerns the work of disagreement at a single table.
- The TIME Framework. The four forms read here as a family of lenses.
- Quadriform. The coordination state the (+T+I+M)^+N test protects.
Sources
- Keller, M., and Haupt, M. (2026, September 25). Zoom dialogue on the ladder of inference, lens and self-talk. Unpublished conversation. Michael Keller is co-founder of Ecology of Design in Human Systems: https://www.ecologyofdesigninhumansystems.com/about-us/about-michael/. The dialogue in which lens and self-talk were named as additions to the ladder.
- Argyris, C. (1990). Overcoming Organizational Defenses: Facilitating Organizational Learning. Allyn and Bacon. The ladder of inference and the reasoning habits that seal beliefs against evidence.
- Senge, P., Kleiner, A., Roberts, C., Ross, R., and Smith, B. (1994). The Fifth Discipline Fieldbook. Currency Doubleday. The widely used version of the ladder, with the left-hand column exercise.
- Ronfeldt, D. (1996). *Tribes, Institutions, Markets, Networks*. RAND P-7967. The four forms read here as lenses.
- Ross, L., and Ward, A. (1996). “Naive realism in everyday life: Implications for social conflict and misunderstanding.” In E. S. Reed, E. Turiel and T. Brown (Eds.), Values and Knowledge (pp. 103-135). Erlbaum. Source entry.
- Ross, L., and Stillinger, C. (1991). “Barriers to conflict resolution.” Negotiation Journal, 7(4), 389-404.
- Pettigrew, T. F. (1979). “The ultimate attribution error: Extending Allport’s cognitive analysis of prejudice.” Personality and Social Psychology Bulletin, 5(4), 461-476.
- Sherman, D. K., Nelson, L. D., and Ross, L. D. (2003). “Naive realism and affirmative action: Adversaries are more similar than they think.” Basic and Applied Social Psychology, 25(4), 275-289. Source entry.
- Nasie, M., Bar-Tal, D., Pliskin, R., Nahhas, E., and Halperin, E. (2014). “Overcoming the barrier of narrative adherence in conflicts through awareness of the psychological bias of naive realism.” Personality and Social Psychology Bulletin, 40(11), 1543-1556. Source entry.
- Kahan, D. M., Braman, D., Gastil, J., Slovic, P., and Mertz, C. K. (2007). “Culture and identity-protective cognition: Explaining the white-male effect in risk perception.” Journal of Empirical Legal Studies, 4(3), 465-505. The closest empirical treatment of a lens: worldview shaping which facts people accept. Source entry.
- Thompson, M., Ellis, R., and Wildavsky, A. (1990). Cultural Theory. Westview Press. The source of the fatalist position.
Provenance
Written 26 September 2026 from a working conversation that placed the ladder of inference in the context of TIME and the BioConomy, then added lens and self-talk following a Zoom dialogue with Michael Keller on 25 September 2026. The bias findings from Sherman, Nelson and Ross (2003) and Nasie et al. (2014) were checked against the primary sources during the companion research brief.