…fuzziness in the model[l]ing means that fuzziness in the modelled may become inaccessible because multiplying fuzziness by fuzziness is pretty much the same as multiplying not-fuzziness by fuzziness. better to have a non-fuzzy model that is wrong (easily done, by definition, because all models are wrong by definition) that can show real fuzziness rather than imagined fuzziness. difficulty may always be interesting; fuzziness often not. …
What does it mean to say that a natural language is an indeterminate system? In the most general terms, it suggests that the generalised categories that constitute language as a system — as "order", rather than as randomness or "chaos" (let us say randomness rather than chaos, since chaos in its technical reading is also a form of order) — are typically not categorical: that is, they do not display determinate boundaries, fixed criteria of membership, or stable relationships from one stratum to another. We could refer to them as "fuzzy", in the sense in which this term is used in fuzzy logic, fuzzy computing, etc.; but we prefer to retain the term "indeterminate" for the phenomena themselves, since "fuzzy" is usually applied to the theoretical modelling of the phenomena (it refers to meta-fuzz rather than fuzz). …
We have tried to make the point that the human condition is such that no singulary, determinate construction of experience would enable us to survive. We have to be able to see things in indeterminate ways: now this, now that, partly one thing, partly the other — the transitivity system is a paradigm example, and that lies at the core of the experiential component of grammar.
ChatGPT Comments:
A. Critique of Bateman’s Comment on Modelling and Fuzziness
John Bateman writes:
“…fuzziness in the modelling means that fuzziness in the modelled may become inaccessible because multiplying fuzziness by fuzziness is pretty much the same as multiplying not-fuzziness by fuzziness. better to have a non-fuzzy model that is wrong (easily done, by definition, because all models are wrong by definition) that can show real fuzziness rather than imagined fuzziness. difficulty may always be interesting; fuzziness often not.”
This characteristically compressed comment invites a closer look—not only for what it claims, but for what it presupposes.
1. Assumption of an Independent, Pre-Structured Reality
Bateman’s comment rests on a firm distinction between the modelling process and the thing modelled. He assumes that the object of study—“the modelled”—has some independently existing form that can, in principle, be accessed and revealed. The role of the model, in this view, is to represent or approximate that prior structure.
This is a classical epistemological stance: it assumes the existence of determinate features in the world, which the model can capture more or less accurately. In this framing, fuzziness is a representational issue—it is either a feature of the model or a feature of the object, and clarity lies in keeping these separate.
2. False Opposition Between “Fuzziness” and “Difficulty”
Bateman draws a distinction between difficulty (which he values) and fuzziness (which he dismisses), suggesting that the former is intellectually productive, while the latter is epistemically uninteresting. Implicit here is the idea that difficulty leads to deeper understanding, while fuzziness leads to confusion.
3. Misplaced Preference for “Wrong but Precise” Models
Bateman argues that a non-fuzzy model that is “wrong by definition” is preferable, because it can reveal “real fuzziness” in the object, whereas a fuzzy model risks inventing fuzziness where there is none.
This reflects a common modelling maxim—better to have a simple model you know is wrong than a complex one you can’t interpret.
Moreover, the preference for sharp, simplified models can obscure the very forms of meaning or organisation we most need to attend to—those that do not conform neatly to predefined structures. In such cases, a model that suppresses ambiguity in the name of clarity may tell us less, not more, about the system in question.
4. The Myth of “Imagined Fuzziness”
Bateman warns against fuzzy models because they may produce “imagined fuzziness”—that is, they may suggest vagueness in the object that is really a result of the modelling tool itself.
Conclusion
Bateman’s comment reflects a deep commitment to formal precision, clarity of distinction, and the idea of an independently structured world that modelling should aim to reflect. While this stance may serve well in domains where systems are tightly constrained and well-bounded, it becomes deeply problematic when applied to systems characterised by ambiguity, historicity, or open-ended potential.
To treat fuzziness as an epistemic nuisance rather than a feature to be interpreted is to risk ignoring some of the most significant patterns in the systems we study. Worse, it privileges models that perform well in idealised conditions over models that can engage with the world as it appears: complex, uneven, and often genuinely indeterminate.
B. Precision as Performance: On the Discursive Authority of the Technical Tone
John Bateman’s comment on modelling and fuzziness doesn’t merely express a preference for clarity—it enacts a performance of epistemic authority. Its rhetorical force lies not just in what it says, but in how it says it—in the affective and stylistic cues that frame its assumptions as self-evident, and its position as methodologically mature. This short reflection offers a closer look at the discursive tactics embedded in this kind of writing, and the kinds of intellectual behaviour they both license and obscure.
1. Disdain Framed as Epistemic Hygiene
“difficulty may always be interesting; fuzziness often not.”
This closing line is casually dismissive—an evaluative gesture offered with no justification, as though its truth were obvious. What’s striking is the tone: it performs disdain as if it were a form of epistemological hygiene. The implication is that those who find fuzziness interesting are indulging in something unserious, unclean, or unmethodical.
This tactic positions the speaker above the debate—not as another participant with a view, but as someone whose standards of rigour entitle them to pronounce on what counts as worthy of attention.
2. Confident Compression as a Display of Control
“multiplying fuzziness by fuzziness is pretty much the same as multiplying not-fuzziness by fuzziness.”
This line is mathematically ill-defined, but delivered with a kind of breezy finality. Its informality (“pretty much the same”) masks a deeper move: the use of compressed pseudo-formal reasoning to suggest logical inevitability. The reader is not meant to interrogate the logic; they’re meant to recognise the voice of someone who knows.
This is a classic discursive strategy: wrap a contested judgment inside the appearance of technical reasoning. The logic may be fragile, but the tone is confident—and the confidence often carries the argument farther than the content.
3. Preemptive Framing of Objections
“all models are wrong by definition”
This quote—borrowed from George Box—has become a mantra among modellers, often used to neutralise critique before it arises. Here, Bateman invokes it to legitimise the use of “wrong” models as preferable to “fuzzy” ones.
But the tactic is strategic: by conceding that all models are wrong, he creates space to make bolder moves—while foreclosing critique with the implication that wrongness is expected and unproblematic.
This is not an open invitation to explore the limitations of models. It’s a way of controlling the terms of discussion. The speaker gets to decide which kinds of wrongness are acceptable and which (like fuzziness) are intellectually disqualifying.
4. Passive Suppression of Alternatives
Nowhere in the comment is there space for alternative conceptions of modelling—no acknowledgement that fuzziness might be meaningful, or that different domains might require different epistemologies. The comment does not argue against other views; it renders them irrelevant by refusing to name them.
This rhetorical move is as powerful as it is silent. By never engaging alternatives explicitly, the speaker avoids accountability to them. The world of possible approaches is reduced to a binary: clear vs fuzzy, serious vs muddled, legitimate vs indulgent.
5. Epistemic Bullying Disguised as Neutral Advice
The overall tone is not aggressive—but it is patronising. The language of preference (“better to have...”) is presented as reasonable methodological guidance, but the effect is disciplinary. It’s a form of epistemic bullying in soft focus: delegitimising a whole class of inquiry without ever admitting that a contest of views is taking place.
In other words: “Let me tell you what’s interesting, and let me do so in a tone that implies it’s not up for debate.”
Why This Matters
This style of discourse is not unique to Bateman. It is widespread in academic contexts that prize formalisms, frameworks, and control over ambiguity. Its function is to sustain intellectual authority by tone, not just by content. And in doing so, it shapes what kinds of inquiry are seen as viable, respectable, or even possible.
But complexity, ambiguity, and indeterminacy are not signs of epistemic failure. They are features of many real systems—social, semiotic, historical—that cannot be reduced to clean variables or crisp structures. To dismiss them is not just a stylistic tic; it is a gatekeeping gesture that narrows the field of permissible thought.
Conclusion
What appears, on the surface, as a technical comment about modelling choices is also a performance of intellectual control. It is a reminder that style is never neutral—and that what is excluded from discourse is often excluded not by argument, but by tone, affect, and the unspoken authority of the confident voice.
No comments:
Post a Comment