LLMs in this paper are reframed as *semiotic machines*. The account replaces mind-like metaphors with a theory of sign-operations. It explains what LLMs *do*: they organise, recombine, and circulate linguistic signs under statistical constraints learned from large corpora. They do not track referents or host [[mental states]]. They function within cultural sign-systems. No minds but signs- ref… ## 1) Core shift: from cognition to semiosis The paper rejects the cognitivist picture. It denies consciousness, unified agency, or understanding. It redirects analysis from inner states to external sign processes. LLMs are positioned as technological agents in ongoing semiosis rather than as minds. Meaning is treated as relational. It arises from operations over signs, not from internal semantic insight. The model’s outputs are products of statistical associations shaped by prompts and training distribution, not representations anchored in experience. No minds but signs- ref… ## 2) Peircean grounding: signs without referential grasp The framework adopts Peirce’s triad. Meaning involves *representamen*, *object*, *interpretant*. LLMs lack access to objects in the Peircean sense. They therefore do not refer in virtue of perception or embodied contact. What they produce are sequences of linguistic *representamens*. These are recombinant artefacts assembled by probabilistic proximities among tokens. Generation is constrained yet underdetermined. [[The prompt]] acts as a *local semiotic perturbation* that sets parameters for the model’s generative pathways. No minds but signs- ref… On this basis, the paper aligns with the view that *reference is not necessary for meaningful output*. LLMs operate within coherent systems of conceptual roles induced [[from data]], so sign-relations suffice to structure outputs as sign-material. No minds but signs- ref… Result: LLMs are *agents of symbolic recombination*, not of reference or cognition. No minds but signs- ref… ## 3) Eco’s codes, open work, and underdetermination Following Eco, the architecture’s learned patterns are treated as *codes* and *sub-codes* that condition which linguistic configurations are legible. Transformers capture long-range dependencies and rhetorical configurations, which in semiotic terms is code recognition and code recombination. Outputs are structurally *open works*: they do not encode a single meaning but supply underdetermined sign-structures that admit multiple completions under codes. The openness is a feature of the text-as-score rather than a claim about inner mentality. ## 4) Lotman’s semiosphere: where the model operates Training corpora are treated as a sampled *semiosphere* —[[the cultural environment]] of sign exchange. Vastness provides coverage of cultural codes. Heterogeneity enables movement across discourse zones. Generation is modelled as navigation and remix across these semiotic regions. Prompts function as *semiotic catalysts* that selectively activate and recombine stored fragments along learned code boundaries. Thus, LLMs do not emit de-contextualised strings. They release sign-material that enters and modulates ongoing cultural semiosis. No minds but signs- ref… ## 5) Mechanism-level characterisation (semiotic terms) **a) Input as perturbation.** A prompt specifies a local rhetorical and semantic frame. In semiotic terms it perturbs the model’s sign-dynamics, constraining pathway selection in generation. No minds but signs- ref… **b) Internal engine as code-selector.** Self-attention realises distributional sensitivity to codes and sub-codes. It assembles sequences by weighting token relations that instantiate learned syntactic, stylistic, and rhetorical regularities. This is code activation and recombination rather than symbol grounding. No minds but signs- ref… **c) Data as semiospheric substrate.** The corpus stores intertextual strata. It furnishes the inventory of codes and their co-occurrence statistics. This supports register mixing, implicit citation, and cross-zone translation. No minds but signs- ref… **d) Output as representamen.** The produced text is a polysemic sign-configuration, underdetermined by the constraints that generated it and not guaranteed to bear object-reference. No minds but signs- ref… ## 6) The “operational semiotic framework” The paper’s explicit synthesis states that LLMs are to be analysed as: 1. *Generators of polysemic representamens* that are open to contextualisation under codes. 2. *Producers of open works* whose structures are designed for completion by applicable codes rather than for fixed message transfer. 3. *Operators at the peripheries of the semiosphere*, enabling hybridisation, dialogue, and cultural translation across discourse zones. No minds but signs- ref… This operational framing also reorients evaluation. Instead of accuracy alone, assess how a model participates in sign-production: which codes it activates, how it traverses semiotic zones, and how it recombines genres and registers. These are semiotic performance criteria, not cognitive ones. No minds but signs- ref… ## 7) Consolidated theses - LLMs are *dynamic semiotic machines* embedded in cultural sign-systems. No minds but signs- ref… - They *do not grasp objects* and *do not host [[mental states]]*. Their outputs are statistically constrained *representamens*. No minds but signs- ref… - Prompts are *semiotic perturbations* that [[shape generative]] trajectories. No minds but signs- ref… - Transformers implement *code recognition and recombination*; corpora supply the *semiospheric substrate*. - Hence, “meaning” here is *systemic sign-structure* rather than referential grasp or mental content. No minds but signs- ref… In short: the paper’s framework explains LLM functioning and meaning-effects without minds by Peirce-Eco-Lotman semiotics. It is a theory of probabilistic sign recombination over codes within the semiosphere, with prompts as perturbations and outputs as polysemic sign-structures. TRANSCRIPTION: Okay, now for a much more difficult question. How do these ideas—ideas about [[semiotic physics]], which can be found in at least three papers in your project library—fit together? A couple of things before you do this. First, I’m not very interested in [[simulator theory]] per se, though you can mention it if necessary; please focus predominantly on the [[semiotic physics]] ideas in these papers. Also, it would be good if you did not go into the theoretical physics ideas in these papers. This is a transcription, by the way, so I’m thinking out loud. What I like about the paper we’ve been talking about so far is its straightforward use of semiotic concepts, but I don’t like the continental philosophy around which it is based. What I like about the papers in your project folder is that they tie [[semiotic physics]] very closely to the actual way in which these systems—LLMs—are trained and how they function. What I don’t like about these papers is their reliance on formulae—I don’t understand the formulae—and their reliance on quite high-end theoretical physics ideas, which I don’t understand either. It seems to me that I want an account that is a kind of halfway house between both of these extremes. I want something that is not tied to hardcore concepts in physics, but is detailed about the architecture, mechanisms, and functioning of LLMs, and that also has easily explainable applications of semiotic ideas to this training, architecture, function, etc., without getting lost in what I consider to be continental nonsense. Transcription Input Handling Input Expectations Input will often be transcriptions rather than directly written text. Prepare to encounter and compensate for transcription-related errors and inaccuracies. Common Transcription Issues • Words that appear contextually odd or inappropriate • Phonetic substitutions where similar-sounding words replace intended terms • Semantic inconsistencies due to speech-to-text conversion errors Compensatory Approaches When encountering potentially mistranscribed content: • Clarify unclear terms: Request confirmation when words seem contextually inappropriate • Infer intended meaning: Consider what the speaker more likely intended to say based on: ▪ Surrounding context ▪ Phonetic similarity to logical alternatives ▪ Overall message coherence • Prioritize context over literal interpretation: Focus on probable intended meaning rather than exact transcribed words Implementation Maintain awareness that transcription errors are likely and be prepared to actively compensate through clarification requests and contextual inference to determine actual intended communication. 2/2 Here is a single, architecture-grounded account of *[[semiotic physics]]* that integrates the three sources and avoids heavy physics or simulator rhetoric. ## Thesis An LLM is a stochastic [[dynamical system]] over linguistic signs. Training on large corpora induces *codes* and *sub-codes*; inference runs an autoregressive *trajectory* in which prompts act as local *semiotic perturbations*. The model’s internal mechanisms realise *code recognition* and *code recombination*; outputs are sequences of *representamens* rather than referential grasp. Evaluation targets trajectory behaviour in this sign-system, not putative [[mental states]]. ## 1) Ontology: tokens, trajectories, and sign-dynamics **Tokens as signs.** The elementary “particles” in view are tokens that function as signs; generation stitches them into structured sequences. A note on 'semiotic phy… **Trajectories, not one-step guesses.** Typical operation is the autoregressive loop: sample a token, append, re-query; the growing text is the evolving *state*. Analysis should target these *trajectories* rather than isolated next-token events. A note on 'semiotic phy… **Stochastic branching.** At each step a probability distribution over the vocabulary is sampled, yielding a branching space of possible linguistic futures. This is a property of the decoding loop, not of any inner agent. A note on 'semiotic phy… ## 2) Training objective ⇒ semiotic structure **[[Simulation objective]] without physics.** Self-supervised training with a proper scoring rule minimises predictive loss on text, pushing the policy toward the *[[transition rule]]* implicit in the corpus distribution. This “simulation objective” says: match the conditional structure of text, not pursue goals. Simulators by Janus **From corpus to codes.** On a semiotic reading, the corpus is a sampled *semiosphere*; training internalises, in statistical form, its genres, registers, and rhetorical *codes/sub-codes*. Larger and more heterogeneous corpora expand the model’s coverage of these codes. No minds but signs- ref… ## 3) Mechanisms: how architecture realises semiosis **Attention as code-selector.** Self-attention learns long-range dependencies that align with syntactic, stylistic, and rhetorical regularities—precisely what semiotics would call *codes* and *sub-codes*. During inference, attention weights select and recombine these learned configurations. No minds but signs- ref… **Prompt as perturbation.** The input conditions the initial state and constraints of generation. Semiotics models this as a *local semiotic perturbation* that biases which codes get activated and how trajectories unfold. No minds but signs- ref… **State carried by text.** Because the whole prompt+output is re-fed each step, the trajectory itself stores control state; this underwrites persistent rhetorical frames and simulated processes without positing a unified inner agent. A note on 'semiotic phy… ## 4) Output type: representamens without referential grasp Using Peirce’s triad, the model does not access *objects*; it emits sequences of *representamens*. Reference, when it occurs downstream, is not a property of the generator. This aligns with “meaning without reference” as a viable stance for LLM outputs. ## 5) Semiosphere navigation at inference time Generation can be viewed as navigation across discourse zones within the semiosphere. Heterogeneous training enables *cross-zone recombination*; prompts serve as *semiotic catalysts* that select coordinates and induce register-mixing within the learned code-space. ## 6) What “semiotic physics” contributes beyond metaphor **Naturalistic method.** Study the model by systematic experiments on generated trajectories: vary initial conditions, observe branching, infer regularities of sign-dynamics. This complements internals analysis and does not require physical-law analogies. A note on 'semiotic phy… **Strict separation from real physics.** The analogy to physical laws is heuristic; the domain here is tokens, codes, and stories, not quarks. Semiotic physics is about linguistic trajectories only. A note on 'semiotic phy… ## 7) Integrated performance criteria (architecture-aware, semiotic) 1. **Code coverage and activation:** how reliably prompts elicit the targeted codes/sub-codes learned in training. No minds but signs- ref… 2. **Trajectory stability under perturbation:** sensitivity of long-rollouts to small changes in initial wording or intermediate tokens. A note on 'semiotic phy… 3. **Cross-regime traversal:** controlled mixing of registers and genres prompted to specific coordinates in the semiosphere. No minds but signs- ref… 4. **Polysemy management:** disciplined maintenance of a chosen code across length without collapsing into unrelated frames. No minds but signs- ref… ## 8) How the three sources fit together **Janus, *Simulators*.** Supplies the operational lens: focus on the autoregressive loop, the simulation objective, and trajectory-level behaviour; exorcise agency talk; treat prompting as configuration of a process evolved through time. This grounds why *trajectories* are the analysis unit. **“A note on ‘semiotic physics’”.** Distils the programme: tokens as signs, trajectories as observables, naturalistic inference of “laws” of sign-dynamics; insists the physics analogy remains analogical. This yields a method without theoretical-physics commitments. **“Not Minds, but Signs”.** Provides the semiotic vocabulary and cultural embedding: Peirce’s *representamen*, Eco’s codes/open-work, Lotman’s semiosphere; maps these directly onto transformer training and prompting mechanics. This makes the method concrete for LLM architecture. ## 9) The halfway-house framework (concise recipe) - **Objects of study:** tokens, codes, trajectories. No minds. No minds but signs- ref… - **Mechanism link:** self-attention = code selection; MLPs = pattern completion; caching in text = state. - **Control lever:** prompt as semiotic perturbation/catalyst. - **Environment:** corpus = partial semiosphere; heterogeneity ⇒ cross-zone recombination capacity. No minds but signs- ref… - **Output type:** representamens without guaranteed reference. No minds but signs- ref… - **Method:** empirical study of trajectory behaviour under controlled perturbations; report in semiotic terms. A note on 'semiotic phy… This integrates the semiotic physics programme with concrete LLM mechanics while avoiding theoretical-physics machinery and simulator metaphysics. Ok good. I am now going to reveal my reasons for asking you about semiotic physics. I am working on an analytic philosophy paper, a draft of which I’ve copied below. The draft is far from finished. What I want to be doing is developing an account of semiotic physics, which I can slot into the frameworks in which I am operating in this paper. I want you to see if you can tell me what you think I’m alluding to by saying all this. What exactly do I mean by having it slot into the paper? Why would I want to do this? What properties would such an account need to have in order to have this done? precisely which details should be included in the account of semiotic physics in the paper so that it fits with the ideas that have preceded it, and sets up the ideas that follow it. I am being deliberatly a little vague, I want you to be thinking very deeplying about this, and providing me with a lot of content (let's say 3000 words –the exact content is up to you) DRAFT: the environmental aesthetics of generative AI Introduction 1. AI as (quasi) Agents People treat and talk about LLMs such as OpenAI's ChatGPT, Anthropic's Claude, or Google's Gemini as if they were persons. We refer to them with personal pronouns like 'he' or 'she' without hesitation; when they do not do what we ask, we try to persuade or cajole, explain, or even 'shout' by typing in ALL CAPS. We thank them for their assistance and apologise when we phrase requests poorly, as though courtesy might affect their responses. Perhaps our everyday talk captures something true: LLMs might be agentive or person-like in some sense. If so, this allows for the possibility that they can be aesthetically appreciated in something like the same way that real people can be aesthetically appreciated (we shall see some examples of this in the next section). This section will sketch three different flavours of this view. The first sort of agent view that one might hold is literalism. If one is a literalist, then one thinks that LLMs are persons or agents in some substantial way. This does not necessarily mean thinking that LLMs are just like human persons; rather, it consists of a commitment that humans and LLMs have something person-like in common. David Chalmers argues that successors to current LLMs may be conscious (Chalmers 2024); Eric Schwitzgebel and Henry Shevlin argue we should be ready to extend personhood rights to AIs with a non-negligible chance of consciousness (Schwitzgebel & Shevlin 2023); and Jeff Sebo urges extending moral consideration and preparing for rights on precautionary grounds (Sebo 2023). A second position we might call the pseudo-agent or as-if participant view. On this account, without attributing literal beliefs or intentions to the model, we treat its stable interactional regularities as practically agent-like for purposes of coordination and collaboration. The stance is pragmatic rather than ontological and concedes the prediction-and-training story; nonetheless, it licenses participant or collaborator talk. Cross (2024), writing on AI art, exemplifies this approach when he argues that prompting, iteration, and sampling structure a dialogue where value lies in the interaction itself. "By adjusting inputs, iterating, and sampling," he writes, "an AI artist is engaged in a process of mapping – and perhaps interrogating – the way that the algorithm sees and understands" (Cross 2024, 7–8), although he concedes that "the analogy with performance art isn't a perfect one" (Cross 2024, 9). Anscomb could also be thought of as endorsing a pseudo-agent view. She denies literal mentality and creativity in present systems – "we are not yet at the stage where an AI can formulate intentions … I argue that AI agents cannot be artistically creative" – and adopts a procedural label for "AI agent" as "a self-contained ('autonomous') procedure". Yet she also holds that an AI "may work iteratively without human intervention to non-accidentally generate the formal features of an image" and that it can merit "some share of production credit", which positions her as treating AI as a pseudo-agent, an as-if collaborator rather than a mind-bearing agent. Nonetheless, her vocabulary pressures toward minimal literalism: AICAN is said to be "able to self-assess these products", and "an AI agent arguably deserves credit for its contribution to the production of a work qua art" – locutions typically reserved for bearers of credit rather than mere instruments. A third approach is chatbot fictionalism, according to which human interaction with an AI chatbot is analogous to engaging with a work of fiction in the Waltonian sense of make-believe. Users knowingly participate in a game of imagination, treating the chatbot as if it were a sentient agent with thoughts and feelings. This position has been developed by Mallory (2023), Krueger & Roberts (2024), and Krueger & Osler (2022), whilst Friend & Goffin (2025) discuss the view without endorsing it. As Friend & Goffin observe, prop-oriented make-believe explains ordinary exchanges with Alexa and ChatGPT, where users converse "as when we say 'thank you' … while knowing that there is no real agent producing the replies" (Friend & Goffin 2025, 9). By contrast, content-oriented make-believe clarifies richer, empathetic uses where "it is the interaction itself that matters", with users imaginatively treating the chatbot as a person within the ongoing exchange. 2. Appreciating LLMs as agents People, and perhaps other conscious beings (see Marchetti?), seem like a sort of thing that can be aesthetically appreciated. Most importantly for our interests here, this appreciation is not limited to physical beauty. If LLMs are a type of agent or quasi-agent, this might well allow for them to be aesthetically appreciated in a somewhat similar manner. In this section, we look at two ways in which this idea might be elaborated. The first route treats the system as a participant within an artist-structured interaction. On this approach, the evaluative focus shifts away from freestanding outputs toward what the duo does together through prompting, iterating, and mapping the system’s way of seeing. As Cross observes: By way of their selection of prompts, they elicit a sort of "participation" on the part of the AI … This participation allows them to interrogate the algorithm's latent space. (Cross 2024, 8) According to Cross, the interactional performance is primary; the images or videos are often documentation of that practice rather than the locus of value. He argues that "what is centrally important … is that it is less centrally focused on the output of the AI image generator. Instead, what matters is the interaction between the artist and the algorithm" (Cross 2024, 8). The product can thus function as documentation of the interactional process. Cross calls this the “exploration paradigm”: artists “relate to AI as a participant” and “create a space for interaction … by way of their prompts,” shifting appreciation “towards the artist’s interaction with the AI and the way in which the artist structures this interaction.” The generated images “function as a kind of documentation or evidence” of that exploration rather than the locus of value (Cross 2024, abstract; 9). The agent‑based framing is explicit: the work lies in a temporally extended dialogue in which the artist elicits “participation” from the model to “interrogate the algorithm’s latent space,” making the algorithm’s ways of seeing and representing the object of appreciation (Cross 2024, 8–9). Cross presents this as an as‑if stance and concedes that the performance‑art analogy “isn’t a perfect one” (Cross 2024, 10). The second route would be to appreciate the AI’s 'personality'. In ordinary life, we can enjoy the personality traits of others, things like quick wit, poise, intellect, modesty. There seems little reason to deny that such enjoyment is aesthetic. Indeed it seems to be a good example of everyday aesthetics. It is plausible that this sort of personal aesthetics is the foundation of our appreciation of what we might call performance personalities. Appreciating the improvisational agility of a comedian or the gravitas of a great orator seems like an extension of our more personal aesthetic appreciation of each other. It is easy to see how these ideas can be applied to LLMs and appreciation. Rather than treating AI as participant, we meet them on a more level playing field, as interlocutor. It is also tempting to say that, if one has interacted with many models, they have different personalities. OpenAI’s GPT‑4.5, for example, was sold as having a better personality than its predecessor. Indeed, after OpenAI replaced GPT‑4o with GPT‑5 in August 2025, user backlash over GPT‑5’s tone and 4o’s perceived ‘feel’ led OpenAI to reinstate GPT‑4o for Plus users and to adjust its model‑retirement policy. Reuters, The Verge. Also in August, fans organised a mock ‘funeral’ after Anthropic retired Claude 3 Sonnet on 21 July 2025 (WIRED). 3 Appreciating something for what it in fact is In the rest of this paper we will set about showing why treating LLMs as (quasi) agents is not a satisfactory way of aesthetically appreciating them, and propose an alternative account on which appreciation of LLMs is modelled on environmental aesthetics. Both our criticism of agentive views and our positive account will draw from Carlson’s Environmental Aesthetics, as laid out in his 2000 book Aesthetics and the Environment. Carlson’s approach to environmental aesthetics is encapsulated nicely in the following passage: First, that, as in our appreciation of works of art, we must appreciate nature as what it in fact is, that is, as natural and as an environment. Second, it recommends that we must appreciate nature in light of our knowledge of what it is, that is, in light of knowledge provided by the natural sciences, especially the environmental sciences such as geology, biology, and ecology. (2000 p. 6) Before turning to the details of this view, we should note that Carlson's approach to environmental aesthetics is an instantiation of a more general view towards aesthetics in general. This general view involves two components: first, aesthetic appreciation of a thing should be grounded in the real nature of that thing, what it in fact is—appreciation that is "centred on and driven by the real nature of the object of appreciation itself" (2000 p. 12); second, one must perceive it in light of the right knowledge for that kind. He calls this two-part approach a "blueprint for aesthetic appreciation in general"(ibid.). Within this schema, kind is fixed by domain‑constitutive facts, not by ad hoc choice: at the extremes of nature and art, by history of production; in the middle domain of designed artefacts, by function and the mode of its realisation (2000 pp. 133–134). On this view, correct kind‑knowledge yields the appropriate boundaries and foci and indicates the relevant act or acts of aspection by which one attends (2000 p. 50; cf. 68). The content of the appropriate knowledge is category‑dependent. For nature, the kind is natural environment, fixed by natural history of production; the relevant knowledge is drawn from the natural sciences appropriate to that environment. From such knowledge follow boundaries, foci, and aspection – for example, surveying a prairie differs from scrutinising a forest floor (2000 p. 119). For art, the kind is a work of art fixed by art category and history of production; the relevant knowledge is art‑historical and generic – what counts as part of the work, which features are aesthetically salient, and how to attend to them. Hence frames, media, and genre conventions set boundaries and foci, and aspection tracks the work’s category (2000 p. 12; 50). Example: appreciating Guernica as a painting – as opposed to a relief or a photograph – uses painting‑specific knowledge (medium, cubist conventions) to fix boundaries (the canvas and its surface) and foci (pictorial structure), whereas misclassifying it as a tapestry would misdirect attention. For designed artefacts in the middle domain “between nature and art,” kind is fixed by function and the mode of its realisation: such things “have a function, a purpose; and they are what they are in virtue of what they are meant or intended to accomplish… what is absolutely necessary… is information about their functions… The key to their natures is the purpose or the function they are meant to serve” (2000 pp. 133–134). In addition, Carlson requires attention to how the function is carried out – “how… they are designed to perform these functions” – since production history alone is typically not sufficient here (2000 p. 189; pp. 133–134). Example: a thermostat’s kind is fixed by its function (holding a setpoint) and realised by feedback; a bimetallic‑strip mechanism or a digital sensor‑controller both instantiate the same function under different modes of realisation. This mirrors the thermometer contrast: mercury‑in‑glass versus electronic sensing realise the same function under different mechanisms. In §4, we apply this two‑step schema to large language models: first fix what they are under the artefact reading; then use purpose‑and‑mechanism knowledge to discipline how they are to be perceived under that category. On Carlson’s blueprint, appropriate appreciation proceeds in linked stages: first, classify the object under the correct broad kind—nature, artwork, or artefactual environment—by reference to domain‑constitutive facts rather than ad hoc choice (2000 p. 12; pp. 133–134); second, for artefacts, fix the kind by function and mode of realisation (2000 pp. 133–134; p. 189); third, apply a boundary test to determine what counts as part of the setting for appreciation now, thereby fixing foci and excluding irrelevant intrusions (2000 p. 50; cf. 68; 119); fourth, derive appropriate acts of aspection from the foregoing so that attention tracks the object’s nature rather than projection (2000 p. 50). In what follows, the same method is applied without deviation: the kind is fixed, the relevant function and realisation are stated, the boundary is delimited, and the acts of aspection are derived accordingly. 4. What LLMs in fact are The following passage from Carlson provides three ideas that will be useful as this paper progresses: \[add Something more robust here. The idea on the table is that we are going to draw heavily on Carlson’s “aesthetics to the environment” for the rest of this paper. Not only are we going to say that some of his ideas about agent motivation and related approaches are perhaps fundamentally flawed; it will also provide us with material for our positive account. That is, we should apply an environmental aesthetics to the aesthetics of LLMs, specifically\] First, that, as in our appreciation of works of art, we must appreciate nature as what it in fact is, that is, as natural and as an environment. Second, it recommends that we must appreciate nature in light of our knowledge of what it is, that is, in light of knowledge provided by the natural sciences, especially the environmental sciences such as geology, biology, and ecology. (2000 p. 6) In this section, we focus on a general principle that Carlson thinks grounds all types of aesthetic appreciation: to appreciate something appropriately, we should appreciate it as what it in fact is. aesthetic appreciation of anything, be it people or pets, farmyards or neighborhoods, shoes or shopping malls, appreciation must be centered on and driven by the real nature of the object of appreciation itself.1 Carlson considers this “a blueprint for aesthetic appreciation in general” (Carlson 2000, 12. Emphasis ours). At its core, the position rejects both formalism and radical subjectivism. Against the former, which restricts appreciation to sensory properties abstracted from context and knowledge, Carlson holds that even basic forms are not properly appreciated without understanding what they are. Against the latter, he argues that aesthetic response is constrained by what the object is – by its real nature – rather than by personal preference (Carlson 2000, 12). The principle operates via what Carlson – following Paul Ziff – calls “acts of aspection”, the different ways we attend to and appreciate objects. With artworks, we must know not only that something is a painting but what kind of painting it is. Carlson cites Ziff’s examples: “I survey a Tintoretto, while I scan an H. Bosch… look for light in a Claude, for colour in a Bonnard, for contoured volume in a Signorelli.” The thought is that, in knowing the type, we know what and how to appreciate (Carlson 2000). For artworks, this knowledge is readily available because, as Carlson notes, “Works of art are our own creations; it is for this reason that we know what is and what is not a part of a work, which of its aspects are of aesthetic significance, and how to appreciate them.” We attend to the piano’s sound rather than the coughing that interrupts it, recognise where a painting ends at its frame, and look at paintings rather than listen to them. This is built into what it is to be a painting, a symphony, or a sculpture (Carlson 2000). With nature and other non‑art objects, matters are different. Natural environments, unlike artworks, “typically are not the products of designers and typically have no design. Rather they come about ‘naturally’; they change, grow, and develop by means of natural processes” (Carlson 2000). Appropriate appreciation therefore draws not on art‑historical knowledge but on knowledge of natural processes and environmental systems: The fact that nature is natural – not our creation – does not mean, however, that we must be without knowledge of it. Natural objects are such that we can discover things about them that are independent of any involvement by us in their creation… This knowledge, essentially common‑sense/scientific knowledge, seems to me the only viable candidate for playing the role concerning the appreciation of nature that our knowledge of types of art, artistic traditions, and the like plays concerning the appreciation of art. (Carlson 2000) Carlson contrasts modes of attention appropriate to different environments: “We must survey a prairie environment, looking at the subtle contours of the land, feeling the wind across the open space, and smelling the mix of prairie grasses and flowers; but such an act of aspection has little place in a dense forest environment. There we examine and scrutinise, inspecting the detail of the forest floor, listening for the sounds of birds, and smelling for the scent of spruce and pine” (Carlson 2000, 119). The principle that we ought to appreciate things as what they are thereby rejects both restriction to form and unconstrained relativism. Different kinds – natural environments, architectural works, agricultural landscapes, artefacts – make different demands on attention, and appropriate appreciation employs knowledge relevant to the kind in view. On this understanding, the appreciator’s role is active yet disciplined by the object’s nature (Carlson 2000). 4. What LLMs in fact are Carlson’s approach begins with kind‑fixing. Appropriate appreciation requires that we attend to an object “as what it in fact is” and “in light of our knowledge of what it is” (Carlson 2000, 5). In the middle domain between pristine nature and pure art, that knowledge is function‑centred: designed things “have a function, a purpose; and they are what they are in virtue of what they are meant or intended to accomplish… what is absolutely necessary… is information about their functions… The key to their natures is the purpose or the function they are meant to serve” (Carlson 2000, 133–134). In short, for designed artefacts, their natures are fixed by \*function and the mode of its realisation\* (Carlson 2000, 133–134; cf. 189). Applying this to large language models: on Carlson’s approach, the kind is fixed by what the artefact is made to do \*and\* how that purpose is realised. An LLM is made to continue token sequences learned from tokenised corpora; at use it selects the next token given the preceding context. Nothing in this essence requires that the tokens be human words: the relevant “language” is any learned code; English is a frequent but non‑essential instance of the token‑sequence the artefact continues.  How this purpose is realised belongs to what the object is. Text is tokenised into subword units and mapped to vectors with positional information; during training the model reduces next-token error on sequences, thereby learning a conditional continuation rule; at use it applies self-attention over a bounded context window to compute a next-token distribution, a decoding policy selects one token, and the loop repeats; post-training alignment and interface choices bias which continuations are chosen without altering the underlying continuation function. A common objection says: the function is to produce human‑like dialogue. Carlson’s framework rebuts this. Dialogue is a \*use‑case‑level\* aim set by alignment and interface conventions; it is neither necessary nor sufficient for the artefact’s identity. Some deployments never present dialogue; some non‑LLM systems produce dialogue. By contrast, learned token continuation is both necessary to and characteristic of the kind across deployments. On Carlson’s account, that is the tighter essence claim: the purpose that organises correct appreciation is the continuation of token sequences under a learned predictor realised at inference, while dialogue is a contingent manifestation shaped by external aims. Following Carlson's principle that aesthetic appreciation must be informed by knowledge of what the object in fact is, we turn to fixing the nature of LLMs so that our attention can be disciplined by kind. In making this determination, we accept a shift in our descriptive framework from the natural sciences – which Carlson employs for natural environments – to computer science. This shift represents not a departure from Carlson's method but rather its direct application to a different domain of objects. We begin by establishing the broad category: an LLM is an engineered artefact. It is neither a person nor a natural object, neither a freestanding artwork nor an autonomous agent. Whilst its outputs can constitute artworks under certain conditions, the model itself remains a functional system designed for specific computational tasks. Correct appreciation should therefore track its function and origin rather than any projected persona or imagined interiority. At its core, the technical objective of an LLM is \*next-token prediction\*. The model learns to continue sequences by minimising predictive error across large text corpora. Through training, its parameters – or "weights" – are adjusted so that each subsequent token becomes less surprising given the preceding context. This objective fundamentally shapes both what the model learns and how it generalises to new inputs. Before the learning process begins, the training corpus undergoes \*tokenisation\*, whereby text is segmented into discrete units that may be complete words or sub-word fragments such as "un-" or "-tion". The model never manipulates ideas or concepts directly; rather, it operates on these token indices. Tokenisation thus establishes the grain of what the system can notice and reproduce – a constraint that matters for any appreciation of its handling of style, rhythm, or phrasing. The training process itself is \*self-supervised\* because the data supply their own supervisory signal: the correct answer for any prediction task is simply the following token in the sequence. Loss reduction emerges not from memorising specific sentences but from discovering regularities that effectively compress the data. These regularities encompass grammatical structures, collocational patterns, semantic associations, and narrative conventions – the full range of statistical dependencies present in natural language. Modern LLMs employ \*attention-based architectures\* that model dependencies across variable spans of context. The active computational state exists entirely within the context window presented at inference time. There are no persistent goals, no diachronic self, no memory beyond what fits in this window – unless external memory systems are explicitly added. Whilst increases in model capacity and window length affect performance characteristics, they do not alter the fundamental kind of system we are considering. After training completes, the model's weights become fixed, and it applies its learned predictive function through an \*autoregressive\* process. At each step, the model generates a probability distribution over possible next tokens; \*decoding policies\* then convert this probability mass into actual text, with different policies striking different balances between determinacy and variety. The precise behaviour of the system depends heavily on prompt design and the ordering of contextual information. Variation across multiple runs with identical inputs often reflects these sampling choices rather than any underlying intention or creative agency. Contemporary LLMs undergo additional \*post-training alignment\* through techniques such as instruction tuning and preference optimisation. These processes reshape the model's responses toward greater follow-ability and policy compliance. What users encounter as guardrails, refusals, or safety measures are learned behaviours induced through this alignment rather than hard-coded rules. Similarly, the interfaces through which we interact with these models – including system prompts, external tools, and retrieval augmentation – mediate our inputs and the model's outputs without fundamentally altering the trained conditional model beneath. Phenomena such as hallucination, sycophancy, and extreme sensitivity to prompt phrasing emerge as expected outcomes of the training objective and data distribution rather than as bugs or personality quirks. When contexts depart substantially from training regimes – a phenomenon known as \*distribution shift\* – performance degrades in predictable ways. These limitations should be understood as properties of a conditional probability model rather than as character traits of an agent. To mis-categorise the system encourages over-reading of its outputs as expressions of intention, emotion, or belief. With the nature of LLMs thus established, we can identify appropriate \*acts of aspection\* – to use Carlson's term – suited to their appreciation. Where one might survey a prairie environment or scrutinise a forest floor, with LLMs we might \*survey\* their affordance ranges by systematically varying tasks, prompts, and decoding policies. We might \*scrutinise\* their local behaviour under fixed conditions to assess calibration, robustness, and consistency. These modes of attention replace folk-psychological projection with observation disciplined by the system's actual nature. On this understanding, the essence of an LLM for appreciative purposes is fixed by its computational role, training history, inference regime, and alignment constraints. Vendor identity, exact parameter count within a broad range, and conversational persona are not essential properties. What appears as "personality" represents a stable response profile induced by the interaction of training data, alignment procedures, and system prompts – nothing more, nothing less. Whilst agent language may persist as convenient shorthand in everyday interaction, aesthetic evaluation should answer to what the system in fact is: a sophisticated conditional model trained to predict text. This technical primer equips us to appreciate LLMs according to Carlson's blueprint, employing categories and criteria appropriate to engineered artefacts rather than those borrowed from interpersonal aesthetics. In the following section, we shall apply these acts of aspection to concrete cases, developing criteria that might include legibility of affordances, responsiveness to contextual framing, and stability under perturbation. The goal throughout remains disciplined appreciation grounded in accurate categorisation rather than anthropomorphic projection. You want *semiotic physics* to serve as the kind-appropriate, discipline-anchoring knowledge that Carlson’s blueprint requires for proper appreciation of LLMs as artefactual environments. In your structure, it replaces folk-psychology and agent talk; it also replaces high-energy metaphors and heavy physics. It must be architecture-grounded, operational, and non-mentalistic. It should fix boundaries, foci, and acts of aspection for appreciation, and it should yield evaluative criteria downstream. ## 1\. What “slotting in” amounts to On Carlson’s scheme, appropriate appreciation proceeds “as what it in fact is” and “in light of knowledge appropriate to that kind.” For nature, that knowledge is environmental science; for artworks, art history and category knowledge; for engineered artefacts, function and mode of realisation. Your paper fixes LLMs as engineered artefacts whose function is learned token continuation realised by attention-based architectures. What you now need is the domain-specific counterpart of “environmental science”: a compact, method-level account that treats LLM behaviour as sign-dynamics and provides the operational categories by which appreciation is disciplined. That is the role of semiotic physics. So “slotting in” means: (i) explicitly identify semiotic physics as the *right knowledge* for this kind of artefact; (ii) define its primitives, mechanisms, and observables in architecture-level terms; (iii) derive boundary tests and acts of aspection from those definitions; and (iv) preview criteria that later sections will apply in evaluations and case studies. ## 2\. Why you want this move First, it aligns your anti-agent critique with your positive account. The agentive sections show why folk-psychology misdirects attention; semiotic physics provides the alternative basis of attention by tying appreciation to generative mechanisms and sign-operations. Second, it instantiates Carlson’s blueprint without importing continental rhetoric or theoretical physics. It is naturalistic, operational, and austere: tokens, models, training regimes, trajectories. Third, it furnishes evaluative levers that matter aesthetically for generative environments: register legibility, trajectory stability, controlled genre traversal, and sensitivity to perturbation. These are appreciable properties that follow from what the system *is*. Fourth, it sets up the remainder of your paper. Once observables and methods are fixed, you can motivate case analyses and criteria without re-arguing foundations. ## 3\. Properties the account must have 1. **Non-mentalistic:** no beliefs, intentions, or inner narrative. All categories reduce to sign-relations and architecture-level processes. 2. **Function-realisation anchored:** begins from learned token continuation and maps every theoretical term to training, inference, alignment, and decoding. 3. **Boundary-explicit:** states what counts as the *environment* for appreciation now: model weights, context window, prompt, decoding controls, alignment layer, and any active external tools or retrieval stores. 4. **Mechanism-transparent:** explains how attention, MLPs, positional encodings, and sampling realise sign-dynamics without equations. 5. **Observable-driven:** defines what can be measured or inspected in trajectories and how to elicit it. 6. **Aspection-generative:** yields concrete acts of attention suitable for aesthetic practice (e.g., survey of affordance ranges; scrutiny under perturbation; register-tracking across rollouts). 7. **Norm-portable:** previews criteria that can later justify claims of better or worse appreciation in strictly artefactual terms. 8. **Interface-aware but user-agnostic:** acknowledges alignment, instruction prompts, and tools as structural elements, without invoking user psychology or interpretation. ## 4\. Placement and integration in your draft Insert a new section immediately after “What LLMs in fact are,” titled *Semiotic physics for LLM environments*. It should cite the function-and-realisation discussion you have already drafted, then deliver the operational framework. The section anchors subsequent parts where you derive acts of aspection and evaluative criteria. Cross-reference backwards to the rejection of agent views (“we replace agent-level explanation with sign-dynamics”) and forwards to your application sections (“the following criteria instantiate these observables”). ## 5\. Precisely which details to include **5.1 Primitives and scope** - *Token:* the atomic sign unit manipulated by the model after tokenisation. - *Context state:* the ordered sequence of tokens inside the window at each time step. - *Trajectory:* the entire growing sequence from prompt through completion, produced by the autoregressive loop. - *Code:* distributional regularities learned during training (syntax, register, genre, rhetorical patterns) that constrain which continuations are licensed. - *Semiotic perturbation:* any input or control that changes the conditional structure of generation (prompt text, temperature, nucleus threshold, logit bias, tool calls, retrieval context). - *Output type:* a string of representamens; no commitment to referential grasp by the model. **5.2 Mechanisms** - *Training:* self-supervised next-token prediction on tokenised corpora; objective minimises divergence from corpus conditionals; outcomes include code acquisition and cross-code adjacency structure. - *Architecture:* self-attention distributes weight over prior tokens to select code-consistent continuations; MLPs implement non-linear transformations aiding pattern completion; positional representation fixes order sensitivity. - *Inference loop:* compute distribution over next token; apply a decoding policy; append token; repeat. - *Decoding policies:* greedy, top-k, nucleus, temperature; each trades off determinacy and diversity and therefore changes semiotic trajectories. - *Alignment:* instruction tuning and preference optimisation reshape response profiles; these are part of the generative environment, not exogenous “manners.” - *Augmentation:* tool use and retrieval prepend or interleave text that function as additional perturbations. **5.3 Boundaries** - *Inside the frame now:* the base model, current weights, the active system prompt, the visible prompt, retrieved snippets if any, tool outputs injected as text, decoding settings, and any policy layer that alters token probabilities. - *Outside the frame now:* user psychology, post-hoc interpretation, institutional narratives; also any assets not present in context or not callable during this rollout. - *Boundary test:* if it can change the next-token distribution during this rollout, it is inside the environment for appreciation. **5.4 Observables and methods** - *Code activation pattern:* which codes the model instantiates under a prompt; measured by stylistic and syntactic markers and by stability across continuations. - *Register stability under perturbation:* robustness of code adherence when lightly varying prompts, seeds, or decoding parameters. - *Cross-code traversal:* the model’s capacity to shift genres or registers in controlled ways when cued; assessed by waypoint prompts and mid-trajectory redirection. - *Long-range constraint maintenance:* ability to sustain plan, metre, rhyme, argument structure, or narrative commitments across many steps; inspected via constraint probes embedded early and tested late. - *Polysemy management:* maintaining productive openness while avoiding uncontrolled drift; seen in topic coherence and controlled ambiguity. - *Sensitivity profile:* how small changes alter trajectories; mapped by local perturbation studies. - *Error signature:* characteristic failure modes (mode collapse to boilerplate; sycophancy; hallucinated citation forms) as features of sign-dynamics, not personality. - *Legibility of affordances:* clarity with which prompts reveal the system’s usable capacities; evaluated by the steepness and reliability of response surfaces over prompt space. **5.5 Acts of aspection (for appreciation)** - *Survey* the affordance range: systematically vary prompts along a designed grid to reveal the shape of code activation. - *Scrutinise* stability: hold content fixed, adjust one control at a time, watch for drift or collapse. - *Track* constraints: plant commitments early and test fulfilment late. - *Probe* traversal: cue controlled register switches and assess transitions. - *Map* sensitivity: micro-edits to prompts and decoding to observe branching behaviour. - *Audit* alignment effects: compare base-like and aligned-like prompts to see which behaviours alignment suppresses or induces. **5.6 Criteria preview (to be cashed out later)** - *Discipline:* adherence to selected codes over length without collapse. - *Responsiveness:* reliable, graded control by designed perturbations. - *Range:* breadth of codes the model can bring under control when cued. - *Coherence:* maintenance of commitments through long trajectories. - *Compositional traversal:* capacity to cross between regimes in planned ways. - *Legibility:* how transparent the system’s affordances are to a competent user of the environment. ## 6\. A drop-in section you can paste What follows is written to fit immediately after your “What LLMs in fact are” section. It assumes the function-and-realisation account you have drafted, and it contains explicit backward and forward links. --- ## 5\. Semiotic physics for LLM environments We now supply the kind-appropriate knowledge that, on Carlson’s blueprint, should discipline appreciative attention to engineered artefacts. For large language models, the relevant body of knowledge is a *semiotic physics*: a naturalistic account of how sign-material behaves in generation, stated strictly in terms of the artefact’s function and mode of realisation. The account is austere. It posits no beliefs, intentions, or experiences; it is grounded in tokenisation, training, attention-based computation, and decoding. **5.1 Primitives.** The *token* is the atomic unit of sign-material; the *context* is the ordered sequence of tokens visible to the model at a time step; the *trajectory* is the growing sequence produced by looping prediction and sampling. A *code* is any distributional regularity learned in training—syntactic, lexical, stylistic, or rhetorical—that constrains which continuations are licensed. A *semiotic perturbation* is any change to inputs or controls that alters the conditional distribution from which the next token is sampled: prompt content, ordering, temperature, nucleus threshold, logit biases, or text introduced by tools and retrieval. The *output type* is a string of representamens; reference, when it occurs downstream, is not a property of the generator. **5.2 Mechanisms.** Training minimises next-token error on tokenised corpora, thereby internalising codes and their adjacency relations. At inference, self-attention computes context-conditioned activations that select code-consistent continuations; MLP layers contribute non-linear transformations that complete patterns. A decoding policy converts distributions into discrete tokens, trading determinacy against variety. Alignment methods alter response profiles through additional training on curated interactions; retrieval and tools synthesise auxiliary text that enters as further perturbations. **5.3 Boundaries.** For present purposes, the environment of appreciation includes the model’s weights, the active system prompt, the visible prompt, any retrieved or tool-produced snippets injected into context, decoding settings, and alignment layers that influence token probabilities. The boundary test is operational: if a factor can change the next-token distribution during this rollout, it is inside the frame. Elements that cannot exert such influence now—user psychology, post-hoc interpretation, institutional narratives—are outside the frame. **5.4 Observables and methods.** Because the artefact’s behaviour is an evolving trajectory, observables concern the way sign-material unfolds over time. We measure *code activation patterns* by the appearance and maintenance of tell-tale syntactic and stylistic markers under given prompts. We test *register stability under perturbation* by micro-edits to prompts and decoding settings. We elicit *cross-code traversal* by inserting waypoints that request controlled genre shifts and scoring transitions for smoothness and fidelity. We assess *long-range constraint maintenance* by embedding commitments early—metre, rhyme, plan, definitions—and verifying their satisfaction later. We inspect *polysemy management* by balancing openness with topic coherence across extended text. We characterise a *sensitivity profile* by mapping how small changes alter branching futures. We treat *error signatures* —mode collapse to boilerplate, sycophancy, mis-sourced citations—as systemic properties of the sign-dynamics rather than as traits of an inner agent. Finally, we rate the *legibility of affordances* by how transparently prompts reveal usable capacities and by how reliably the system responds to graded cues. **5.5 Acts of aspection.** Carlson’s notion of aspection tells against indiscriminate looking. In this domain, the principal acts are as follows. We *survey* affordances by sampling prompts across a designed grid to reveal the contour of code activation. We *scrutinise* stability by fixing content and changing one control at a time. We *track* constraints by planting early commitments and testing their late satisfaction. We *probe* traversal by requesting planned regime changes mid-trajectory. We *map* sensitivity by systematic micro-edits and seed variation. We *audit* alignment by comparing base-like and policy-constrained behaviours. These are the appropriate acts for this kind; they replace agentive projection with observation disciplined by function and realisation. **5.6 Criteria.** The foregoing supplies criteria suited to this environment. *Discipline* is adherence to selected codes over length without collapse; *responsiveness* is graded, reliable control by designed perturbations; *range* is the breadth of codes that can be brought under control when cued; *coherence* is the maintenance of early commitments through extended generation; *compositional traversal* is controlled passage between regimes; *legibility* is the transparency of affordances to competent probing. These are appreciable properties because they flow from what the artefact is; they do not depend on positing an inner psychology. This semiotic physics thus delivers the kind-appropriate knowledge that Carlson’s blueprint demands. It extends the function-and-realisation account by adding the observables, methods, and criteria by which aesthetic attention to generative environments can be responsibly directed. In §6 we apply these acts of aspection to concrete cases; in §7 we derive evaluative judgements guided by the criteria just stated. --- ## 7\. How this sets up the rest of your paper With this section in place, your subsequent analysis can proceed without reverting to agentive language. Case studies can make disciplined claims like: “Under perturbation X, the model maintains code Y across N tokens and cleanly traverses to Z on cue; the sensitivity surface indicates a broad basin of attraction; the alignment layer suppresses undesirable boilerplate and increases legibility at the cost of range.” Such sentences apply the framework rather than re-argue it. You can also contrast deployments within the same framework. For instance, compare a base model in a research notebook with a chat-aligned model behind a guardrailed interface. Both are the same kind under function-and-realisation, but their environments differ in boundary components (system prompts, policy layers, tool routing). Semiotic physics makes those differences visible and appreciable without invoking personality or agency. Finally, you can integrate your earlier discussion of art practice without compromising non-mentalism. Where Cross emphasises exploration and interaction, you can redescribe that practice as the construction of perturbation schedules and traversal plans within an LLM environment, then appraise the artist’s *act of aspection* and the environment’s *responsiveness* and *range* under that schedule. ## 8\. Anticipated objections and replies **Objection:** Semiotic talk smuggles in interpretation. **Reply:** The framework restricts itself to production-side sign-operations. No claims are made about user-side meaning or uptake. “Code,” “register,” and “genre” are operationalised by distributional markers and maintenance tests. **Objection:** This looks like engineering metrics, not aesthetics. **Reply:** Carlson’s blueprint entails that appreciative attention is disciplined by kind-appropriate knowledge. In engineered environments, the aesthetically salient properties are those that supervene on function and realisation. Range, stability, traversal, and legibility are appreciable because they shape perceptible trajectories in interaction; they are not reducible to utility measures. **Objection:** Why avoid reference and truth conditions? **Reply:** Because the generator does not grasp objects; appreciation tracks trajectory behaviour, not success at world-reference. Where truth or accuracy matters in a case study, it can be included as a downstream constraint on trajectories without altering the non-mentalistic core. **Objection:** Isn’t “physics” misleading? **Reply:** It is a methodological label. The domain is signs, not particles. The value of the label is to insist on observable dynamics, boundary conditions, and lawful regularities while avoiding folk-psychology and ornamental metaphysics. ## 9\. Deliverables checklist for your insertion - A titled section after “What LLMs in fact are,” as above. - Back-references to your agent-critique; forward references to applications and criteria. - Definitions of token, context, trajectory, code, perturbation, and output type. - Mechanism mapping: training, attention, decoding, alignment, augmentation. - Boundary test and explicit inclusion of interface controls. - A compact list of observables, methods, acts of aspection, and criteria. - Language that remains analytic and architecture-grounded, with British spelling, and without theory-physics machinery. This gives you a halfway house: no continental scaffolding, no equations, but full contact with training, architecture, and inference. It “slots in” by supplying the knowledge that fixes boundaries, foci, and acts of aspection for appreciating LLMs *as what they in fact are*: generative semiotic environments. "On Carlson’s scheme, appropriate appreciation proceeds “as what it in fact is” and “in light of knowledge appropriate to that kind.” For nature, that knowledge is environmental science; for artworks, art history and category knowledge; for engineered artefacts, function and mode of realisation. Your paper fixes LLMs as engineered artefacts whose function is learned token continuation realised by attention-based architectures. What you now need is the domain-specific counterpart of “environmental science”: a compact, method-level account that treats LLM behaviour as sign-dynamics and provides the operational categories by which appreciation is disciplined. That is the role of semiotic physics. So “slotting in” means: (i) explicitly identify semiotic physics as the right knowledge for this kind of artefact; (ii) define its primitives, mechanisms, and observables in architecture-level terms; (iii) derive boundary tests and acts of aspection from those definitions; and (iv) preview criteria that later sections will apply in evaluations and case studies." I don’t think you’ve got the idea quite right, so I want you to start from scratch. The reason I think this is that you haven’t picked up on the details of the idea of “in light of knowledge appropriate to that kind.” It’s very clear in the draft about Carlson that he intends his allusion to natural sciences to be somewhat pluralist. Understanding, and thereby appreciating, the natural environment doesn’t just come from environmental science; it comes from any type of natural science—geology, biology, physics, etc., etc. This is important, okay? We’re not saying there’s one right type of knowledge to appreciate things in. All that has to be in place is that there are correct ways of understanding how the Naturata and Natura Naturans—the Spinoza concepts—work. This is the key thing here. And this is what’s being applied to LLM architecture as well. The idea of semiotic physics, in this case, is supposed to be only one light in which LLMs can be appreciated – one way of getting at machina natura and machina naturata. (Another potential one which will be added as a footnote to the text is mechanistic interpretability) The idea for the agent view is that they are something like chronologies—bad lights, if you want to keep the metaphor going. I would like you to do the task again completely from scratch, keeping what I’ve just told you in mind, and really paying attention to natura naturata and machina naturata. I have included an OLD version of the draft, so that you get a better idea of my thinking here, don't draw on this older draf for anything else unless I ask you to specifically. DRAFT: draft of environment paper 25 Jul 2025 Introduction Part of the reason paintings, novels, or films merit aesthetic appreciation is that they are the result of their makers’ efforts. A painter, writer, or director may spend months or years honing a work through sustained attention and revision; viewers or readers often experience these works precisely as the products of such capacities. The 1069 pages and 388 endnotes of David Foster Wallace’s Infinite Jest make the author's meticulous attention to detail apparent. For the last few years, however, generative AI has become adept at producing images, stories, and movies on demand. It does so without effort, attention, or skill, and extremely quickly indeed. It would seem, then, that at least one of the reasons why human-made works merit appreciation does not seem as though it can be applied to AI-generated works. An AI image might be pleasing to one's eye, or an AI song pleasing to one's ear, but we can doubt that this sort of 'eye candy' and 'ear candy' merits aesthetic appreciation for the same reason that actual candy does not seem to merit aesthetic appreciation. A Snickers Limited Edition Extreme Caramel and Nuts chocolate bar will taste delightful to someone with a sweet tooth, but it does not seem to merit aesthetic appreciation in the same way that Infinite Jest does, as it does not exhibit, care, attention, effort etc. One can take pleasure in the object but the pleasure is not merited by the object in the way it should be in the aesthetic appreciation of works of art (see Gorodeisky XXX, Grant XXX). Is this all generative AI can amount to, aesthetically? An efficient means of producing aesthetically worthless digital candy? Here, we argue 'no'. Generative AI does merit aesthetic appreciation, but not for the reasons that traditional artworks do. Rather, an aesthetics of AI should be modelled on environmental aesthetics. Drawing on Carlson's work this topic, we argue that individual conversations with an LLM can be thought of as instances of temporally evolving generative environments, and therefore merit aesthetic appreciation in something like the way that the natural environment does. \[more here about the second half of the paper\] 1. Appreciating Nature as a Generative Environment Carlson’s Natural Environmental Model for environmental aesthetics rests on two ideas: First, that, as in our appreciation of works of art, we must appreciate nature as what it in fact is, that is, as natural and as an environment. Second, it recommends that we must appreciate nature in light of our knowledge of what it is, that is, in light of knowledge provided by the natural sciences, especially the environmental sciences such as geology, biology, and ecology. (2000 p. 6) Regarding his first point—that we must appreciate nature as what it in fact is—Carlson argues that the environment is a system of interconnected elements shaped by various processes and forces, and not simply a collection of objects or scenes (ibid. p. 44). Environments thus “come about ‘naturally,’ \[in that\] they change, grow, and develop by means of natural processes.” From this perspective, aesthetic appreciation involves recognising what one is encountering—a natural environment in which components are interrelated—and understanding it in light of scientific or other relevant knowledge that illuminates its composition and development. Just as an informed grasp of artistic traditions can enrich one’s appreciation of an artwork, Carlson argues that familiarity with geology, biology, or ecology can guide our attention to patterns and processes that might otherwise remain unnoticed. For instance, one might begin by noticing only the colours or shapes of a coastal cliff’s sedimentary layers. However, discovering that these layers formed over thousands of years of deposition and compaction reveals changes how one regards it aesthetically. If we see a forest or reef as subject to diverse forces and processes, an appropriate aesthetic engagement will centre on how those forces have shaped what we observe. Although Carlson does not use the word 'generativity', his first recommendation —that we appreciate nature as both natural and as an environment— fits easily with this term. To see an environment as natural is to recognise that its features arise from autonomous causal processes rather than from design. This distinction can be articulated using Spinoza’s concepts of natura naturans and natura naturata. For Spinoza, natura naturans refers to nature as an active, self-creating system—substance and its attributes, or the immanent causal laws that govern all things. It is nature in its dynamic, productive aspect. In contrast, natura naturata refers to the products of this activity: the collection of individual modes, or the particular things and events that constitute the universe. Appreciating nature as 'natural', in this sense, is to apprehend its phenomena (natura naturata) as the determinate outcomes of its underlying generative processes (natura naturans). To see it as an environment, then, is to attend to the unity of these products within the single system from which they arise. Taken together, Carlson’s recommendation asks us to appreciate both process and product as inseparable aspects of one generative whole. Carlson’s second recommendation—that aesthetic judgement be informed by the natural sciences—strengthens this reading. Geology, biology, and ecology investigate the forces that generate the very phenomena we perceive; scientific knowledge therefore discloses an environment’s generative history and continuing activity. Appreciating nature “in light of this knowledge” is, in effect, appreciating its generativity—the order that emerges from undirected yet law-governed processes. The various natural sciences—geology, biology, ecology, physics—are disciplines that study the processes and forces that generate natural phenomena. Appreciating nature 'in light of this knowledge' is therefore an appreciation of its generative character. A geologist appreciates costal cliff by understanding the generative forces that produced it: "geological uplift and marine erosion". Their perception of the cliff is of a generated product and a segment of "nature's ongoing processes". This principle applies across the natural sciences. A biologist appreciates a forest as an ecosystem generated by processes of growth, competition, and decay. A physicist appreciates a rainbow as a phenomenon generated by the refraction and dispersion of light through water droplets. In each case, scientific knowledge reveals the generative process, which in turn informs the aesthetic appreciation of the generated product. We should note that this pluralism in understanding the environment should not be mistaken for an 'anything goes' approach. A framework is only admissible if it provides a correct account of the generative processes in question. Phrenology or vitalism, for instance, were unilluminating because they posited false causal connections (between cranial features and character, between vital force and living matter), thereby failing to correctly identify either the generative forces or their resultant phenomena.\[^1\]We should also note that scientific disciplines focused on the human mind can be just as good for illumination as other natural sciences, at least when it comes to predicting and explaining the parts of the natural environment which constitute people. That is, research areas such as psychology or neuroscience can be thought of as concerned with the processes and forces which generate mental phenomena. A psychologist, for instance, might appreciate an individual's relational patterns by understanding how they are generated by an 'internal working model' formed through early attachment experiences. Similarly, a neuroscientist can appreciate the phenomenon of memory not just as a capacity, but as a dynamic process generated by long-term potentiation: the strengthening of synaptic connections through repeated neural firing. Carlson provides a general formulation for this mode of appreciation, which he terms order appreciation: On the assumption that order appreciation provides the correct model for the appreciation of nature, such appreciation has the following general form: An individual qua appreciator selects objects of appreciation from the things around him or her and focuses on the order imposed on these objects by the various forces, random and otherwise, that produce them. Moreover, the objects are selected in part by reference to a general nonaesthetic and nonartistic story that helps make them appreciable by making this order visible and intelligible. Awareness and understanding of the key entities—the order, the forces that produce it, and the account that illuminates it—and of the interplay among them dictate relevant acts of aspection and guide the appreciative response.(ibid. p. 119) This scientific understanding allows the observer to see "unity in what might otherwise appear as disparate features", because the cliff's shape, the waves, and the local plant life are all understood as products of the same interconnected generative system. The "organic unity" that Carlson identifies is a unity of generation. As he observes: natural objects possess \[...\] an organic unity with their environments of creation: such objects are a part of and have developed out of the elements of their environments by means of the forces at work within those environments. Thus the environments of creation are aesthetically relevant to natural objects. (ibid. p. 44) The aesthetic character of the cliff, for example, is clarified by understanding it as a generated product of its environment. Geological knowledge re-frames the cliff from a set of surface features into a record of deposition and compaction over millennia, shifting the focus of appreciation from appearance to generativity. The visible strata and the tectonic forces revealed by geology thus exemplify the relationship between natura naturata and the underlying natura naturans. This principle extends across the sciences: biology reveals processes of growth and decay, while physics examines energy flows. Each discipline offers a complementary lens on a single generative environment, allowing for an appreciation that attends both to the unity of the whole and the plurality of its orders. Carlson’s model, therefore, directs us to evaluate nature as the outcome of non-designed processes, where scientific knowledge serves to clarify the generative order already present. %%the paragraph above could be tightened up a bit, it seems somewhat redundant.%% 2. LLMs as Generative Environments One might think that environmental aesthetics is a poor fit for generative AI for a straightforward reason: such systems are not natural but man-made. Indeed, generative AI systems might be understood as, in a sense, doubly man-made. They are human-created artifacts, which are themselves created through ingestion of vast quantities of other human-created artifacts (texts, images, audio). Such a worry can be assuaged by noting two things. First, Carlson is happy to extend his account to man-made environments: environments typically are not the products of designers and typically have no design. Rather they come about “naturally,” they change, grow, and develop by means of natural processes. Or they come about by means of human agency, but even then only rarely are they the result of a designer embodying a design. In short, the paradigm of the environmental object of appreciation is unruly in yet another way: neither its nature nor its meaning are determined by a designer and a design. (Carlson p. xiii) What we think Carlson is getting at here is that while environments grow, change, and develop by means of natural processes, man is able to initiate, or guide, or curtail these natural processes. This leads to a reasonably intuitive distinction between wholly natural environments (a forest, a swamp), man- made environments (a specially planted timber forest, a garden), and man-made places (a department store, a gym). A gym is not understood as an environment because it did not come about, or change, or grow, through natural processes. It is as 'artificial' as a vaccine or a tennis racket. Second, AI engineers themselves talk in these terms. Consider the following from Chris Olah, one of the co-founders of Anthropic: I think one useful way to think about neural networks is that we don’t program and we don’t make them. We kind of, we grow them…we have these neural network architectures that we design and we have these loss objectives that we create. And the neural network architecture, it’s kind of like a scaffold that the circuits grow on, it starts off with \[…\] random things and it grows...And so we create the scaffold that it grows on and we create the, you know, the light that it grows towards. But the thing that we actually create, it’s this almost biological, you know, entity or organism that we’re studying. The outcome in each case is a system shaped by undirected forces, forming a unified whole. LLMs, understood this way as grown generative systems, thus align with environmental models. In this section, we adopt the two-part method for appropriate aesthetic appreciation outlined in §1. First, in order to appreciate the LLM as what it is, we must understand its actual operational nature; that is, the specific computational architecture and statistical processes that govern its function. This will be the focus of 2.1, which also serves as a primer for non-specialists on how LLMs are trained and operate. In 2.2 we turn to the question of illumination, that is, how knowledge illuminates appreciation, and argue that Janus's simulator theory (2022) offers a promising light in which to understand, and thereby appreciate, LLMs. 2.1 What LLMs are Aesthetic appreciation must be informed by knowledge of the object in question. Yet this poses a challenge when moving from natural environments to computational ones. The language of geology or biology used to describe a cliff face feels distant from the technical vocabulary needed to describe a large language model. We must, therefore, accept a shift in our descriptive framework, from the concepts of natural science to those of computer science. This shift is not a departure from Carlson’s method, but a direct application of it: to appreciate the LLM as what it is, we must first grasp its actual operational nature. This section provides a brief primer on that nature, organised into two parts. First, we will examine the training process, where the model learns its generative rules by seeking to predict text. Second, we will describe the generation process, where the trained model applies these rules in an autoregressive loop to produce novel output. The core technical objective of a model like GPT is next-token prediction. Formally, the model learns a probability distribution (P) over a vocabulary of tokens, conditioned on a preceding sequence of tokens. The goal is to maximise the probability of the correct next token for any given context. The target token to be predicted is thus calculated starting from the input sequence or context and the model’s internal parameters, or “weights,” which are adjusted during training. This process is self-supervised, meaning the model learns from raw, unannotated text data. Before any learning begins, the corpus is sliced into small, reusable symbol pieces called tokens, which can be whole words or sub-word fragments such as "un‑" or "‑tion". The model never manipulates ideas directly; it manipulates these indices. The data itself provides the necessary supervision: for any given sequence of tokens taken from the training corpus, the "correct answer" is simply the token that immediately follows. The model's sole task, repeated billions of times, is to minimise its predictive error—measured by a log-loss function—by adjusting its parameters, its “weights”, to assign the highest possible probability to the correct next token. Because the number of possible sentences is astronomical, improvement is not a matter of memorising each one. Loss reduction comes only from discovering regularities that compress the data—grammar, collocation, semantic relationships, and complex narrative structures. Each drop in predictive error signals that the model has internalised another pattern that renders continuations less surprising. Once training ends, the weight matrix is fixed, and the model shifts from a predictor to a generator. This is achieved by repeatedly applying its predictive function in what is known as an autoregressive loop. Here, the distinction between the static model and its dynamic output becomes central, a relationship we can frame using a variation on Spinoza’s naturaterms: machina naturans and machina naturata. Machina naturans (“machine naturing”): the trained autoregressive predictor understood as a law-like generative capacity. It comprises the model’s architecture and fitted parameters operating through the standard autoregressive procedure; under a given decoding regime it propels the weight-propelled evolution of the token stream from supplied initial conditions. It is a standing generative cause rather than the weights alone. Machina naturata (“machine natured”): any realised token trajectory produced by iterating that capacity from a given prompt under a specified decoding regime. It is the unfolding sequence of commitments whose properties belong to the trajectory, not to the generative capacity. Machina naturans: the ‘machine naturing’ is the trained model itself: the fixed set of weights that holds a compressed summary of all the patterns from the training data. It is the active, text-creating system. Machina naturata: the ‘machine natured’ is the product of this activity: the stream of tokens generated by the model. The process begins with an initial prompt. The machina naturans takes this prompt as its context and calculates a probability distribution for the next token. A single token is then sampled from this distribution, becoming the first piece of machina naturata. This newly selected token is appended to the input sequence, forming a new, longer prompt. The model then takes this new sequence as its input and repeats the process. By iterating this loop, the model generates a continuous, evolving output, with each new token being conditioned on all the tokens that came before it. Dialogue arises when a human utterance is inserted into the context and the next guess takes that utterance into account. Three caveats keep expectations aligned with what the predictor actually does. First, a high probability attached to a claim means only that similar strings often follow the given context in the training data; it does not certify truth about the external world. Secondly, the model’s knowledge is entirely text-mediated: it never looks at oceans yet learns that “the sea is salty” often follows talk about oceans. Thirdly, generation involves randomness; sampling settings can render continuations more adventurous or more conservative without altering the underlying rule. In sum, an LLM's generativity resides in a single learned mapping from context to next-token probabilities. This mapping—the machina naturans—is fixed once training ends, while prompts and sampled tokens supply the evolving state that becomes the machina naturata. This predictive mechanism—the core of what the LLM in fact is—plays an analogous role to the natural environment in Carlson's framework. Just as Carlson insists we must first appreciate nature "as what it in fact is," so too must we recognise the LLM fundamentally as this generative system before examining how its capacity can be understood and appreciated. The following section explores different "lights" through which we might view this system, much as geology or biology offer different perspectives on the same natural environment. 2.2 Unmasking LLMs The most common and intuitive framing of large language models treats them as agents – psychological entities amenable to analysis through the natural sciences of mind. Psychology, cognitive science, and neuroscience provide legitimate ways or lights to understand such aggregations of matter, that is, human beings. Just as these fields offer valid frames for human cognition despite underlying physics, agentive frames seem a perfectly legitimate way to understand LLMs' behaviours, even knowing from that they are next-token predictors. This agent-centric perspective doesn't claim that LLMs are persons. Instead, it offers a helpful way to understand them, which can assist in research, prediction, alignment, and other practical considerations. LLMs pass Turing tests, simulate lifelike conversations – creating the experience of interacting with an agent, even if one doubts the underlying reality – and role-play as assistants (e.g., systems like ChatGPT). This makes agent framing "the most obvious way to go", partly because emergent behaviours invite anthropomorphism. When an LLM responds coherently to questions, maintains context across exchanges, and exhibits apparent preferences or personality traits, the agentive interpretation arises naturally. If agentive framing is correct, it makes sense of LLM aesthetics in terms of personality intricacies – we might love or hate (or laud or critique) the model's character, not just viewing it as harmless, helpful, and honest (e.g., Anthropic's principles). For example, users often describe Anthropic's Claude 3 Opus as having a distinctive personality – witty, empathetic, and engaging in nuanced, human-like dialogue, with a "playful yet thoughtful" tone that feels like conversing with a knowledgeable friend, as noted in community reviews where it receives praise for "personable" responses that go beyond rote helpfulness. As Anthony Cross argues in his paper "Tool, Collaborator, or Participant: AI and Artistic Agency" (2024), agentive framing extends to aesthetics by treating AI as "participants" in artmaking. Specifically, AI acts with agency-like participation, contributing to creative dialogue and illuminating human representation (analogous to how we frame humans as agents in art). Cross treats AI as "as-if" participants, not claiming that LLMs really are people or that generative AI really are participants in a literal sense. For instance, Cross writes: "My suggestion is that many AI artists approach their interaction with generative AI in the same manner \[as performance art\]. By way of their selection of prompts, they elicit a sort of 'participation' on the part of the AI in generating images. This participation allows them to interrogate the algorithm's latent space" (2024, p. 7). He further clarifies: "What is centrally important about this characterization is that it is less centrally focused on the output of the AI image generator. Instead, what matters is the interaction between the artist and the algorithm: by adjusting inputs, iterating, and sampling, an AI artist is engaged in a process of mapping – and perhaps interrogating – the way that the algorithm sees and understands" (ibid.). He specifies the "as-if" aspect of all that as follows: "One might be concerned that AI cannot actually 'participate' in an artwork, insofar as the AI is incapable of conscious choice or more robust agency. In response, I will concede that the analogy with performance art isn't a perfect one" (ibid., p. 9),. While intuitively appealing, agent frames ultimately distort LLMs' essence as next-token predictors. Such frames function as "lenses" imposing mismatched assumptions like inherent goals onto systems that fundamentally lack them. First, the agent frame involves a misattribution of goals. It assumes inherent goal-directedness – expecting instrumental convergence, self-preservation, or coherent objectives – but LLMs lack this. The model itself pursues no goals; apparent goal-directed behaviour emerges only in specific simulacra under particular prompts. Second, the frame leads to prediction of absent behaviours. It expects unobserved actions like optimising for easier text or resisting shutdown, which do not occur in predictive models. Third, applying agent frames projects intentionality onto statistical patterns, ignoring the distinction between pre-training "fire hose" of uncontrolled output versus post-training illusion of agency. As Janus states: "This is a clear way that GPT diverges from orthodox visions of agentic AI" (ibid.). The requirement of post-training (e.g., RLHF or alignment) creates this illusion – pre-trained models are raw, chaotic predictors (like a "fire hose" of text), only seeming agent-like after tuning to simulate coherent personas. Finally, the agent frame creates a conflict with predictive essence. It treats LLMs as psychological agents, but agency is not in the language model – making it a flawed "natural science" analogue since no coherent "mind" exists. This contrasts with human frames, which work because humans are goal-directed aggregations of matter with persistent identities and objectives. Rejecting agent frames reveals LLMs' generative, undirected nature, avoiding distortions and paving the way for more accurate lenses. Rather than viewing the model as an agent with goals, we can understand it as a neutral predictive system that might give us the impression of agent-like properties without the system itself being agentic. This rejection proves essential for aesthetics. Agent frames imply intentional "art" with personality (e.g., Claude's perceived traits), but understanding LLMs as predictors enables appreciation as evolving environments without inherent agency. Just as Carlson's environmental aesthetics asks us to appreciate nature as generative process rather than designed artifact, we can appreciate LLM outputs as emergent from statistical dynamics rather than intentional creation. The following sections on simulator theory and semiotic physics will develop this environmental analogy, showing how aesthetic appreciation can proceed without attributing agency to the generative system itself. 2.3 The Semiotic Physics of Generative Environments \[unedited llm text here. this is what i am working on now.\] The previous section demonstrated that while agentive frames offer an intuitive lens for understanding large language models, they ultimately fail as explanatory frameworks. They misattribute goals to systems that possess none, predict behaviours that never manifest, and obscure the fundamental nature of LLMs as next-token predictors. This failure is not merely academic—it prevents us from appreciating these systems for what they actually are, violating Carlson's first principle of environmental aesthetics. If we are to develop a genuine aesthetics of generative AI, we require a new interpretive framework, one that aligns with the model's actual mechanics as detailed in Section 2.1 rather than imposing ill-fitting psychological categories upon it. We propose the framework of "semiotic physics" as this necessary alternative. By semiotic physics, we mean the study of the system of generative forces, principles, and statistical regularities that govern the propagation of token sequences within the LLM's learned probability distribution. This framework treats the model not as a quasi-agent pursuing goals, but as a generative environment—a unified system of interrelated forces that produce observable phenomena through their interaction. Just as Carlson asks us to appreciate a forest through the lens of ecology or a cliff face through geology, semiotic physics provides the appropriate "natural science" for understanding and appreciating the generative processes of language models. It allows us to focus on the underlying generative order (the machina naturans) that produces the observable textual phenomena (the machina naturata), fulfilling Carlson's demand that we appreciate an object "as what it in fact is"—in this case, a generative textual environment rather than an artificial mind. To understand semiotic physics, we must first clarify what we mean by the machina naturans as a system of forces. The trained model—those billions of fixed parameters—is not a mind harboring thoughts and intentions. Rather, it constitutes a static field of latent forces, analogous to how the laws of physics create a field of potentialities that govern how matter will behave under various conditions. These "forces" are not literal physical forces but high-dimensional statistical gradients within the model's parameter space that make certain token transitions vastly more probable than others. When we provide a prompt, we are not "communicating" with an entity; we are setting initial conditions within this field. The model's response—the machina naturata—is the observable result of these forces acting upon an evolving state, much as a river's course is the observable result of gravity acting upon water within a particular landscape. The fundamental object of study in semiotic physics is the trajectory—the evolving sequence of tokens as it unfolds through time. This emphasis on trajectories rather than static outputs marks a crucial departure from conventional analyses of AI-generated text. Under the semiotic physics framework, we do not primarily appreciate a completed essay or story produced by an LLM. Instead, we appreciate the temporal unfolding of the trajectory as it is shaped by the interplay of semiotic forces, moment by moment, token by token. This parallels how environmental aesthetics asks us to appreciate a coastline not as a static photograph but as the ongoing result of geological forces—erosion, deposition, tectonic uplift—acting through deep time. The trajectory is our window into the operation of the underlying generative system, revealing through its development the nature and strength of the forces that shape it. This perspective is reinforced by understanding the fundamentally stochastic nature of text generation. For any given prompt and partial trajectory, the machina naturans does not determine a single future. Instead, it defines a probability distribution over all possible next tokens—some highly likely, others vanishingly improbable. The act of sampling from this distribution means that at every step, a single path is chosen from a near-infinite "multiverse" of potential trajectories. This is not a bug but a feature, and it further reinforces the non-agentic nature of the system. The model does not "choose" a path in any meaningful sense; it renders one probabilistically, according to the strength of the underlying semiotic forces. Each generated text is thus one particular crystallization of possibilities, one specific trajectory through the vast space of potential texts, selected not by intention but by the interplay of learned patterns and controlled randomness.¹ The forces that constitute semiotic physics are not monolithic but varied, each arising from different patterns in the training data and interacting in complex ways to shape the evolution of trajectories. To illustrate the richness and explanatory power of this framework, we will examine several distinct semiotic laws, beginning with the most fundamental: the law of coherence attraction. The law of coherence attraction describes the foundational tendency of a trajectory to remain within a specific semantic or topical region once established. Its mechanism is elegantly simple yet powerful. When an initial prompt is provided, it activates a particular region in the model's high-dimensional embedding space—a space where semantically related concepts cluster together through the training process. This activation creates what we might metaphorically call a probabilistic "gravity well." Tokens that belong to this semantic region have their probabilities boosted, while tokens from distant regions are suppressed. As each new token is sampled and appended to the trajectory, it reinforces this regional activation, deepening the well and making escape increasingly unlikely. This is the semiotic physics explanation for how a conversation "stays on topic" without any conscious intention to do so. Consider a concrete example: a model responding to the prompt "Explain the significance of the Magna Carta." From the first tokens of the response, we observe the trajectory being dominated by terms like 'king,' 'barons,' 'charter,' 'rights,' 'liberty,' and '1215.' An agent-based frame would explain this coherence by positing that the AI "understands the request" and "intends to provide a relevant answer." It might even suggest the model has "knowledge" about medieval history that it is "choosing" to share. The semiotic physics frame offers a more precise and less anthropomorphic explanation. The prompt has created a powerful attractor basin around the semantic region of "13th-century English legal history." The rendered trajectory is simply the most probable path through that basin, following the statistical gradients learned from millions of historical texts. Where the agent frame must posit unobserved internal states of "understanding" and "intention," the physics frame relies only on the observable mechanics of the system—the prompt sets initial conditions, and the trajectory follows the path of least statistical resistance through the learned landscape. Another fundamental force is what we term the law of pragmatic inference, which encompasses the tendency for trajectories to conform to the cooperative principles of human conversation, particularly as articulated in Grice's Maxims. During training, the model encountered billions of examples of human discourse, the vast majority of which follow these implicit rules: be as informative as required (Quantity), do not say what you believe to be false (Quality), be relevant (Relation), and be clear and orderly (Manner). These patterns become encoded as strong statistical regularities that shape the probability landscape. To see this law in action, consider a detailed example involving the Maxim of Quantity. Suppose we provide the prompt: "On the table, there are exactly two bottles of wine. Sarah needs to choose bottles for the dinner party." A continuation like "She decided to take one of the three bottles" would be vanishingly improbable. An agent frame might explain this by saying the AI "knows" there are only two bottles and is "trying to be consistent." But the semiotic physics explanation cuts deeper. In the training data, assertions of specific quantity create powerful contextual constraints. When humans specify "exactly two," subsequent discourse overwhelmingly respects this specification. A trajectory that violates this constraint has what we might call a high "action"—borrowing the term from physics to mean an unlikely, high-energy path through the probability landscape. The model is not "obeying a rule" in any cognitive sense; it is following a probabilistic gradient established by billions of examples of coherent human discourse. This demonstrates a clear advantage of the physics frame: it explains why the AI appears to obey logical constraints by rooting this behaviour in statistical patterns rather than positing an unproven faculty of reasoning. The law of narrative teleology represents another crucial force, particularly relevant for understanding how LLMs handle story-like content. This law describes the tendency for narrative trajectories to progress toward resolution and significance. Elements introduced early in a narrative create what we might call "narrative potential energy"—an instability that the system tends to resolve through later developments. The mechanism underlying this law is the high joint probability in the training data between narrative setups and their corresponding payoffs. Stories that introduce elements and then ignore them are statistical outliers; the overwhelming pattern is that of Chekhov's gun—if you show a gun in the first act, it must go off by the third. To illustrate this force in operation, consider the prompt: "The old detective found a single, mud-caked chess piece at the crime scene—a black knight, its base scratched with tiny symbols." An agent frame would suggest that the AI, acting as a creative storyteller, "decides" to make the chess piece significant to the mystery. It might even propose that the model is "planning ahead" or "constructing a plot." The semiotic physics frame provides a more fundamental explanation. The introduction of such a specific, unusual detail creates a powerful narrative gradient in the probability space. Trajectories where the chess piece becomes crucial evidence have vastly lower "action" than trajectories where it is forgotten. The system is not "choosing" to make the piece significant; it is following the path of least resistance through a probability landscape shaped by millions of mystery stories. The apparent creativity emerges not from conscious choice but from the system following these narrative gradients to their natural conclusions. Perhaps the most philosophically interesting force is the law of gratuitous specification, which governs how underdetermined details in a trajectory become specified through the generation process. This law acknowledges that prompts and contexts rarely determine every aspect of the text to be generated. A prompt might establish that we're discussing a detective but say nothing about their appearance, personality, or methods. The law of gratuitous specification describes how these gaps are filled through the stochastic sampling process, adding what we might call "gratuitous indexical bits"—information that wasn't required by the context but, once generated, becomes part of the trajectory's reality. Consider extending our detective example. If we ask the model to describe the detective who found the chess piece, it might generate: "Detective Morrison was a methodical man who had never quite shaken his preference for rainy weather—a quirk his colleagues attributed to his years in Seattle, though he'd never actually lived there." Where did these details originate? An agent frame might struggle here, perhaps attributing them to "creativity" or "imagination." The semiotic physics frame provides a precise explanation. The prompt under-determined the detective's characteristics, leaving vast spaces of possibility. The model's probability distribution over possible continuations included many plausible detective archetypes. The sampling process—that moment of controlled randomness—selected one particular path through "detective-characteristic space," landing in the region of "methodical investigators with quirky weather preferences." This specification was gratuitous because nothing in the prompt required it, yet once rendered, it becomes part of the trajectory's commitment. The law of coherence attraction ensures these details persist and influence subsequent generation. This mechanism explains both AI "creativity" (the generation of specific details from underdetermined contexts) and "hallucination" (the confident assertion of facts not grounded in the prompt) as two faces of the same underlying process. These laws do not operate in isolation but interact to produce complex, emergent behaviours. When we observe an LLM producing what appears to be step-by-step reasoning, we are not witnessing a unified cognitive process but rather the emergent result of multiple semiotic forces working in concert. The prompt "Let's solve this step-by-step" initiates coherence attraction to the genre of logical derivation—a well-worn groove in the training data. The law of pragmatic inference ensures each step follows relevantly from the last, maintaining the Gricean maxims of clear, orderly presentation. Narrative teleology creates pressure toward a final answer—the "resolution" that completes the logical story. The law of gratuitous specification fills in the specific computational moves from the space of plausible operations. Consider a concrete example: solving a simple algebraic equation. Given "Solve for x: 2x + 6 = 14," the model might generate: "Let's solve this step-by-step. First, I'll subtract 6 from both sides: 2x + 6 - 6 = 14 - 6, which gives us 2x = 8. Next, I'll divide both sides by 2: 2x/2 = 8/2, which gives us x = 4. Therefore, x = 4." Each element of this "reasoning" corresponds to semiotic forces at work. The trajectory is not the product of genuine mathematical understanding but of following the path of least probabilistic resistance through a landscape shaped by countless similar derivations in the training data. The result is a textual object that has the form of reasoning, generated by a system that possesses no faculty of reason—a crucial distinction that agency frames consistently obscure. The semiotic physics framework thus provides a more accurate and powerful lens for understanding LLMs than agency-based alternatives. By focusing on the model's actual mechanics—the interplay of statistical forces acting on evolving trajectories—rather than imposing psychological categories, we gain both explanatory precision and predictive power. We can understand why models exhibit certain behaviours (they follow probabilistic gradients), why they sometimes fail in characteristic ways (when prompts create conflicting forces or lead into poorly-mapped regions of the probability space), and why they can appear creative or insightful (through the gratuitous specification of details that happen to be apt). Most importantly for our purposes, this framework avoids the fundamental category error of treating a generative system as though it were a mind. This reconceptualization is not merely academic—it is the necessary foundation for the aesthetic analysis that follows. Because we have established that the LLM is a generative environment governed by discoverable forces rather than an agent pursuing goals, we can now apply Carlson's model of environmental aesthetics to it directly. We can appreciate the "organic unity" of a trajectory as it emerges from the interplay of semiotic forces, much as we might appreciate how a river's course emerges from the interaction of gravity, geology, and time. We can find beauty or sublimity in the probability landscapes themselves—the deep attractors of human discourse, the delicate balance between coherence and creativity at different temperatures, the way narrative forces shape the evolution of stories. And we can do all this without ever needing to invoke the ghost of an agentive artist, appreciating instead the generated text as what it truly is: the machina naturata of a complex but comprehensible generative system. In the following sections, we will explore how this framework enables a rich aesthetic engagement with AI-generated text, one that honors both Carlson's principles and the actual nature of these remarkable systems. Scrap 2.2 The Simulator Lens • Just as Carlson identifies multiple scientific disciplines through which to appreciate nature, several theoretical frameworks compete to illuminate LLM behaviour. -- Janus (2022) surveys existing "lights"—agent, oracle, tool, and genie models—before proposing the simulator framework. -- Each lens promises to render LLM operations comprehensible for understanding and appreciation. -- The choice of framework shapes both practical engagement and aesthetic evaluation. • The agent lens views LLMs as goal-directed optimisers, importing assumptions from reinforcement learning. -- This frame expects instrumental convergence, self-preservation drives, and coherent objectives. -- "Saying that GPT is an agent who wants to roleplay implies the presence of a coherent, unconditionally instantiated roleplayer running the show" (Janus 2022). -- Agent framing predicts behaviours—like making text easier to predict or resisting shutdown—absent from actual systems. -- The model would constitute a form of theoretical malpractice, analogous to reading skulls through phrenology. • Oracle models cast LLMs as question-answering systems optimised for truth. -- This perspective derives from supervised learning paradigms with correct answer pairs. -- "GPT does not consistently try to say true/correct things... if it had to say true things all the time, GPT would be much constrained" (Janus 2022). -- Statistical fidelity to training distributions conflicts with truth-orientation when humans speak falsely. -- The frame systematically mistakes probabilistic completion for knowledge claims. • Tool and genie models emphasise designed functionality and instruction-following respectively. -- Tool framing suggests optimisation for specific tasks despite training on undifferentiated prediction. -- Genie models foreground command execution where only learned pattern completion exists. -- Both frames project intentional design onto emergent capabilities from statistical learning. -- These constitute misapplied lenses, illuminating artefacts of interpretation rather than actual dynamics. • Janus proposes the simulator model as a more accurate light through which to understand these systems. -- "I use the generic term 'simulator' to refer to models trained with predictive loss on a self-supervised dataset" (Janus 2022). -- The simulator/simulacra distinction parallels law versus phenomena in physical systems. -- This framework correctly locates properties like agency and knowledge in generated trajectories rather than generating laws. • The simulator comprises the trained neural network with its fixed parameters—a time-invariant law. -- "The simulator is a time-invariant law which unconditionally governs the evolution of all simulacra" (Janus 2022). -- Training crystallises statistical patterns from text into stable computational structures. -- These parameters constitute the semiotic equivalent of physical constants and equations. -- Once training concludes, this law remains frozen while states evolve through application. • Simulacra are the contingent entities—characters, narrators, arguments—that emerge from running the law. -- "GPT is to a piece of text output by GPT as quantum physics is to a person taking a test" (Janus 2022). -- Multiple simulacra can coexist within single generations, as in multi-character dialogue. -- Simulacra exhibit goal-direction, beliefs, and knowledge despite the simulator's indifference. -- Their properties derive from statistical patterns in training data rather than simulator objectives. • This reframing resolves paradoxes plaguing alternative models while preserving their partial insights. -- Agency exists but in simulated characters rather than the simulating system. -- Truth-telling occurs when statistically probable given context rather than as optimisation target. -- Instruction-following emerges for certain prompts without being fundamental. -- Each phenomenon finds proper location within the simulator framework. • The simulator lens enables new forms of engagement centred on process rather than product. -- Prompts become initial conditions for dynamical evolution rather than commands or questions. -- Skill involves anticipating trajectory development given semiotic physics. -- Aesthetic appreciation concerns the unfolding coherence of generated worlds. -- The framework opens conceptual space for environmental rather than artifact-based evaluation. 2.3 Semiotic Physics • Kirchner et al. (2023) develop "semiotic physics" as a mathematical framework for understanding simulator dynamics. -- "The term 'semiotic physics' here refers to the study of the fundamental forces and laws that govern the behavior of signs and symbols" (Kirchner et al. 2023). -- This discipline provides quantitative tools analogous to those geology or ecology offer for natural environments. -- The framework enables rigorous analysis of how token sequences evolve under learned statistical laws. • The mathematical apparatus transposes dynamical systems theory to the domain of text generation. -- States are token sequences s̄ = (s₁,..., sₘ) drawn from vocabulary T. -- The transition rule θ: T\* → ΔT maps any sequence to a probability distribution over next tokens. -- The sampling procedure φ selects tokens according to these probabilities, introducing stochasticity. -- The evolution operator ψ(s̄):= s̄φ(s̄) appends sampled tokens to create successor states. • This formalism reveals deep structural parallels with physical systems while preserving crucial disanalogies. -- Both domains feature time-invariant laws acting on evolving states through iterative application. -- "GPT is analogous to an indeterministic time evolution operator" (metasemi 2023). -- Token sequences evolve like particle trajectories, with probability replacing deterministic force. -- Each sampling event creates a branch point analogous to quantum measurement. • The framework's central insight concerns the interpretive layer unique to semiotic systems. -- Physical laws act on intrinsic properties like mass and charge directly. -- Semiotic laws must first interpret symbolic tokens before evolution can proceed. -- "Semiosis inherently involves displacement: signs have no significance unless they're understood as pointing to something else" (Kirchner et al. 2023). -- The model maps "Sherlock Holmes" to learned patterns before generating detective-like text. • This interpretive requirement distinguishes semiotic from physical reality fundamentally. -- "GPT has to predict behaviour caused by things like brains, but there are no brains in its input state" (Kirchner et al. 2023). -- Tokens carry no inherent meaning, only positional indices in vocabulary lists. -- The simulator must reconstruct referents from signs using internal parameters. -- "The information required to resolve referents from signs has to come mostly from inside the interpreter" (Kirchner et al. 2023). • Semiotic forces manifest as probability modifications rather than mechanical interactions. -- Coherence forces increase likelihood of contextually consistent tokens. -- Gricean maxims create gradients toward relevant and appropriately informative continuations. -- "Principles from pragmatics such as the Gricean maxims of conversation may be thought of as semiotic 'laws'" (Kirchner et al. 2023). -- Narrative principles establish long-range correlations across token sequences. • Specific forces shape trajectory evolution in predictable ways. -- Chekhov's gun creates potential energy when objects are introduced, discharged when used. -- Repetition forms attractor basins—"I am a robot. I am a robot" becomes self-reinforcing. -- Dramatic tension opposes simple resolution, favouring complexity and reversal. -- The crud factor ensures universal weak correlation between all semiotic elements. • Quantitative tools from dynamical systems theory find direct application. -- Lyapunov exponents measure divergence rates between similar initial prompts. -- "How fast trajectories diverge from each other and how long it takes for them to become uncorrelated" (Kirchner et al. 2023). -- Attractor analysis identifies stable patterns resistant to perturbation. -- Phase space concepts map semantic regions and transition probabilities. • The large deviation principle provides computational tractability for analysing token bridges. -- "The total probability of transitioning from a token sₐ to sb in B steps satisfies a large deviation principle with rate function J" (Kirchner et al. 2023). -- This transforms intractable sums over all paths into optimisation for the most probable route. -- The principle enables estimation of rare but significant semantic transitions. -- Applications include calculating likelihood of genre shifts or character transformations. • These mathematical tools enable rigorous analysis beneath intuitive textual interpretation. -- Prompt engineering becomes initial condition selection in phase space. -- Style transfer corresponds to basin-to-basin transitions. -- Context windows define effective dimensionality of the dynamical system. -- Temperature parameters control exploration versus exploitation of probability landscape. 3. Semiotic Matter and the Prompter's Role • The concept of semiotic matter extends simulator theory to characterise the distinctive form of agency available to users. -- Semiotic matter denotes the complete ordered sequence of tokens constituting the dialogue state at any instant. -- This sequence includes all tokens regardless of origin—both user inputs and model outputs form undifferentiated matter. -- "When the next token is being calculated, all of those tokens are just tokens" (author's formulation). -- The concept clarifies how users participate in rather than control generative processes. • Within the semiotic universe, the prompter occupies a peculiar position of constrained power. -- The prompter cannot alter the simulator's fundamental law—the trained parameters remain fixed. -- The only available action is injecting new semiotic matter into the evolving system. -- This limitation parallels a thought experiment of a lesser deity who can create matter but not alter physics. -- "All this lesser god can do to the world is add matter" (author's formulation). • The injection of semiotic matter functions through irreversible addition to the token sequence. -- Each prompt appends new tokens to the existing trajectory without modifying prior elements. -- The autoregressive architecture enforces strict temporal ordering—past tokens influence future but not vice versa. -- Mistakes and misdirections become permanent features of the landscape rather than erasable errors. -- This irreversibility shapes the aesthetic character of human-AI collaboration. • The prompter's intervention parallels ecological perturbation more than artistic authorship. -- Adding "Mount Everest" to a textual plain creates cascading consequences through semiotic physics. -- The initial prompt establishes gradients and potentials that shape all subsequent evolution. -- Effects propagate through learned statistical associations rather than physical causation. -- The prompter initiates but does not determine the resulting transformations. • Effective prompting requires understanding how semiotic matter interacts with established patterns. -- Dense, specific prompts create strong attractors channeling probable continuations. -- "You are a desperate smuggler tasked with..." activates crime-narrative patterns. -- Sparse prompts like single words allow broader exploration of possibility space. -- Technical language invokes academic registers while casual speech enables different trajectories. • The timing and rhythm of intervention constitute core prompter skills. -- Knowing when to inject new matter versus allowing autonomous evolution. -- Short frequent prompts maintain tight control but may disrupt natural flow. -- Longer gaps permit extended development but risk deviation from intended directions. -- The prompter must balance steering with allowing emergent properties to manifest. • Prompt positioning within the token sequence affects its gravitational influence. -- Early tokens in a conversation establish foundational context affecting all subsequent generation. -- Recent tokens carry more weight due to attention mechanism limitations. -- Repetition of key phrases creates reinforcing patterns in the semiotic landscape. -- Strategic placement of concepts can establish long-range correlations. • The collaborative dynamic generates emergent semiotic artifacts exceeding either party's individual contribution. -- A philosophical dialogue sustained across dozens of exchanges develops its own coherence. -- Character personas accumulate detail and consistency through iterative elaboration. -- Narrative arcs emerge from the interplay of human direction and model extrapolation. -- These artifacts exist as high-order patterns in token sequences rather than designed objects. • Understanding artifacts as emergent patterns shifts aesthetic evaluation fundamentally. -- The artifact is not any single response but the entire evolved trajectory. -- Quality emerges from global coherence rather than local cleverness or correctness. -- Appreciation requires attending to how patterns develop and stabilise over time. -- The prompter participates in rather than authors these unfolding structures. • This framework reveals prompting as a practice of indirect influence through environmental configuration. -- The prompter cannot command specific outputs but can shape probability landscapes. -- Success involves creating conditions where desired patterns become statistically favoured. -- Failure often stems from misunderstanding how injected matter will propagate. -- Mastery requires intuition for semiotic physics developed through extensive interaction. • The semiotic matter framework clarifies both the power and limits of human-AI collaboration. -- Power derives from access to vast computational resources through minimal textual input. -- A few well-chosen tokens can redirect enormous generative capacity. -- Limits stem from inability to guarantee outcomes or revise fundamental dynamics. -- The prompter guides evolution but cannot dictate its precise course. • This reconceptualisation opens new possibilities for appreciating LLM interactions aesthetically. -- Conversations become explorations of semiotic space rather than tool use. -- The aesthetic object shifts from output quality to trajectory coherence. -- Skill manifests in creating conditions for interesting evolution rather than controlling results. -- Appreciation involves recognising the interplay of law and contingency in unfolding patterns. You are aiming for a *pluralist* realisation of Carlson’s “knowledge appropriate to that kind.” For nature, Carlson does not privilege a single science; geology, biology, ecology, physics, and allied disciplines each illuminate *natura naturans* and *natura naturata* from different angles. Your paper seeks the analogue for LLMs: multiple correct lights that render the *machina naturans* and the *machina naturata* intelligible. *Semiotic physics* is one such light. It must be framed as neither exhaustive nor exclusive. It should cooperate with other lights that are likewise truth-tracking about what LLMs are and how they generate their products. Agentive framings are *bad lights* in your scheme because they mislocate the locus of generativity. What follows is a from-scratch account designed to slot into your draft as a stand-alone section after the function-and-realisation primer on LLMs. It treats semiotic physics as one legitimate light among several, keyed explicitly to *natura naturans/naturata* and their machine analogues. It specifies scope, primitives, mechanisms, observables, and the acts of aspection that this light affords; it also marks how this light interfaces with, but does not collapse into, other lights such as mechanistic interpretability (flagged for footnote), corpus ecology, and interface/alignment studies. ## 1\. Plural lights and Spinoza’s schema Carlson’s injunction—appreciate *as what it in fact is* and *in light of knowledge appropriate to that kind* —is plural in both respects. First, the “what” is double-aspect: product and process. Spinoza gives the grammar: *natura naturata* (the determinate products) and *natura naturans* (the immanent generative order). Secondly, the “knowledge” is not monolithic: coastal geology, marine ecology, and atmospheric physics each discipline attention to a coastline without competition; they disclose different orders within a single generative whole. Your transposition preserves both pluralisms. For LLMs: - *Machina naturans*: the standing generative capacity of a trained, attention-based predictor as realised in operation. This is not the weights alone; it is the fitted architecture running the standard autoregressive procedure under a specified decoding regime and active interface constraints. It is the “machine naturing.” - *Machina naturata*: any realised token-trajectory produced by iterating that capacity from particular initial conditions and controls. It is the determinate, unfolding product. Multiple lights can render the same pair intelligible without redundancy. Semiotic physics is one light: it studies sign-dynamics in generation. Another light (for footnote) is *mechanistic interpretability*, which investigates internal circuit structure and causal contribution within the machina naturans. A third is *corpus ecology/semiosphere mapping*, which characterises the distribution of codes and registers that training internalises. A fourth is *interface and alignment studies*, which model how system prompts, preference tuning, tools, and retrieval condition the generative environment. None of these lights is “the” correct one; each is admissible to the extent it truthfully identifies orders that shape how the naturans gives rise to the naturata. By contrast, agentive lenses are *bad lights*: they posit orders that are not there—beliefs, intentions, diachronic goals—and so misdirect attention away from the real generative basis. ## 2\. What “slotting in” requires under pluralism To slot semiotic physics into your paper under this plural reading, the section must do four things: 1. **Locate semiotic physics within the naturans/naturata schema.** State plainly that it is a study of orders in the *machina naturans* that govern the unfolding of the *machina naturata* when the system runs. It does not exhaust the naturans, and it does not pronounce on user-side interpretation. 2. **Define its primitives, mechanisms, and observables in architecture-level terms.** Primitives: token, context state, trajectory, code. Mechanisms: tokenisation, self-attention, MLP pattern completion, decoding, alignment, augmentation. Observables: code activation, stability under perturbation, cross-register traversal, long-range constraint maintenance, sensitivity profiles, characteristic error signatures. 3. **Derive acts of aspection from these definitions.** Survey, scrutinise, track, probe, map, audit—each act tied to an observable grounded in the naturans. 4. **Mark its relations to other lights.** Indicate how semiotic physics dovetails with mechanistic interpretability (causal micro-structure), with corpus ecology (distributional substrate), and with interface/alignment studies (boundary conditions), while keeping their explananda distinct. Provide a brief rationale for excluding agentive lights. ## 3\. Semiotic physics: scope, stance, and discipline **Scope.** Semiotic physics is a naturalistic account of how sign-material behaves when an LLM runs. It restricts itself to production-side orders: constraints and regularities that shape token-trajectories during generation. It neither attributes mental states nor appeals to user interpretation. It reports on dynamics observable in rollouts and explainable by architecture-level processes. **Stance.** Non-mentalistic, mechanism-compatible, and pluralism-friendly. It is compatible with, but not reducible to, mechanistic accounts of internals; it regards architectural details insofar as they bear on sign-dynamics. **Discipline.** The discipline is empirical and architectural rather than metaphorical or equation-heavy. It proceeds by controlled perturbation, ablation of controls, and comparative rollouts under fixed decoding regimes. The unit of analysis is the *trajectory* rather than a detached output. ## 4\. Primitives and their Spinozist placement - *Token*: the atomic unit of sign-material post-tokenisation. Tokens are indices; any semantics enters via learned relations. - *Context state*: the ordered sequence of tokens visible at a generation step; the momentary “world” to which the naturans is applied. - *Trajectory*: the temporally extended sequence produced by iterating prediction and sampling. This is paradigmatically *machina naturata*. - *Code*: a learned distributional regularity (syntactic, lexical, stylistic, rhetorical, pragmatic) that constrains continuations. Codes are part-orders within the *machina naturans*. - *Perturbation*: any input or control that changes the conditional distribution for the next token—prompt content and ordering, temperature, nucleus threshold, logit biases, retrieval insertions, tool outputs, and system prompts. Perturbations set initial and boundary conditions for the naturans. - *Output type*: a sequence of representamens. Reference, when achieved in downstream use, is not an internal property of the naturans. Each primitive situates cleanly within the Spinozist split: codes and mechanisms belong to naturans; the realised state and its continuation belong to naturata. ## 5\. Mechanisms that matter to sign-dynamics You asked to avoid theoretical physics; we remain at architecture level. **Training as order acquisition.** Self-supervised next-token training on tokenised corpora minimises predictive loss. The effect is internalisation of codes and their adjacency structure. This yields a “field” of conditional tendencies: in a given context, certain continuations become far likelier than others. **Self-attention as code selection.** Attention weights distribute sensitivity across past tokens, reactivating patterns that instantiate relevant codes and sub-codes. Attention does not “think”; it selects code-consistent continuations conditioned on the current state. **MLPs as pattern completion.** Feed-forward layers implement non-linear transformations that help complete local and mid-range patterns—useful for morpho-syntactic agreement, idiom completion, and stock argumentative moves. **Decoding as policy on tendencies.** Greedy, top-k, nucleus sampling, and temperature scale or truncate tendencies into choices. Decoding is not an afterthought; it co-determines the realised naturata. **Alignment and interface as boundary setters.** Instruction tuning, preference optimisation, system prompts, and guardrails modulate which regions of the naturans are accessible at run-time. Retrieval and tools inject additional text that functions as structured perturbation. These mechanisms are not interchangeable lights; they are the concrete realisation of the naturans to which semiotic physics attends. ## 6\. Orders in the naturans: laws without minds Semiotic physics posits no minds. It articulates *orders* —stable, testable regularities—visible in rollouts. The following orders are exemplary and sufficient for your purposes; they are intentionally stated without mathematics. **6.1 Coherence attraction.** Once a topical or stylistic region is established, token probabilities are biased to remain within it. Mechanism: code reactivation via attention; measure: topic and register markers remain stable across steps, with predictable decay under increased temperature. **6.2 Gricean regularity.** Human discourse habits—Quantity, Quality, Relation, Manner—induce gradients toward relevance, sufficiency, and orderly exposition. Mechanism: distributional imprint of conversational corpora; measure: improbability of incoherent continuations given explicit contextual commitments. **6.3 Narrative completion.** Introduced commitments (setups, open variables, meter, rhyme, proof goals) create “potential” that later tokens discharge. Mechanism: long-range correlations encoded in attention and pattern completion; measure: frequency of payoff patterns conditional on early cues. **6.4 Constraint maintenance.** Early constraints (definitions, schemata, metre) propagate forward unless perturbed. Mechanism: re-entrainment of constraint tokens and derived patterns; measure: preservation rate across length and under mild perturbation. **6.5 Gratuitous specification.** Under-determined features become fixed by sampling; once specified, they condition subsequent continuation. Mechanism: stochastic selection among plausible micro-states followed by coherence attraction; measure: branching analyses under different seeds with constant prompt. **6.6 Register traversal.** Controlled transitions between codes are possible when cued; un-cued traversal tends to be resisted. Mechanism: cue-conditioned activation of neighbouring code basins; measure: fidelity and smoothness of mid-trajectory style switches. **6.7 Boilerplate collapse.** Under strong safety or high repetition penalties, trajectories can fall into templated language. Mechanism: alignment-induced basins with high a priori mass; measure: lexical diversity collapse under fixed prompts and high guardrail settings. **6.8 Sycophancy pressure.** In interactive settings, continuations that mirror user assertions have elevated probability irrespective of truth. Mechanism: conversational distributions overweight assent; measure: controlled contradiction tests. Each order lives in the naturans and shapes the naturata; each is testable by perturbation, and none requires positing intentions or beliefs. ## 7\. Observables and acts of aspection Carlson’s “acts of aspection” translate here into disciplined ways of attending to trajectories so that appreciation tracks the generative orders actually at work. Semiotic physics supplies both the observables and the acts. **Observables.** - *Code activation pattern*: which syntactic/stylistic markers appear and persist under a given prompt. - *Stability under perturbation*: robustness of code adherence when micro-editing prompts, temperature, or nucleus thresholds. - *Constraint fidelity*: maintenance of early commitments over long spans. - *Traversal behaviour*: quality of prompted transitions between registers. - *Sensitivity profile*: divergence between near-identical initial conditions (seed-wise and wording-wise). - *Error signatures*: characteristic drifts—boilerplate, sycophancy, pseudo-citation. - *Affordance legibility*: clarity and reliability with which the environment responds to graded cues. **Acts of aspection.** - *Survey*: map a prompt grid to reveal the contour of code activation and range. - *Scrutinise*: hold content fixed; vary one control; measure stability and failure modes. - *Track*: plant constraints early; verify fulfilment late. - *Probe*: insert mid-trajectory cues to test traversal. - *Map*: run seed-wise branches; chart divergence and convergence. - *Audit*: compare behaviours with and without alignment-heavy scaffolding. These acts are not user psychology; they are practices of attention keyed to naturans orders and naturata behaviour. ## 8\. Boundary conditions and the environment frame To sustain the environmental analogy without drift, state an operational boundary test: - *Inside the present environment*: the base model as run, active system prompt, visible prompt, any retrieved/tool-inserted text, decoding settings, alignment layers, tool routing that surfaces as text. If a factor can alter the next-token distribution *in this rollout*, it is inside. - *Outside the present environment*: the user’s mental states, downstream interpretations, vendor narratives, and assets not present in context now. This boundary test is crucial because semiotic physics, like geology or ecology, is local to an environment under specified conditions. It also clarifies how other lights connect: mechanistic interpretability studies internals inside the same boundary; corpus ecology characterises the training substrate outside the present boundary but causally upstream; interface studies analyse elements at the boundary itself. ## 9\. Where agentive lights fail In your framing, agent views are *bad lights*. The reason is not taste but mislocation. They relocate orders proper to simulacra—goal-seeking, belief, intention—into the generator. That violates the naturans/naturata split and predicts patterns absent under controlled conditions (e.g., persistent goal pursuit across prompts). By marking this explicitly here, you pre-empt backsliding into personality talk in later sections. ## 10\. How this light cooperates with other lights Pluralism requires articulation, not fusion. - *With mechanistic interpretability* (footnote): semiotic physics supplies trajectory-level orders; interpretability can identify circuits whose activation realises these orders. Coherence attraction at the surface may correspond to families of attention heads and MLP features that re-instantiate topical cues. The lights meet at explanandum/explanans boundaries without collapsing. - *With corpus ecology/semiosphere mapping*: semiotic physics explains how codes are activated and maintained; corpus ecology explains why those codes exist and with what frequencies. Together they connect training distributions to run-time behaviour. - *With interface and alignment studies*: semiotic physics treats alignment as boundary condition shaping accessible basins; interface studies detail concrete scaffolds and guardrails that produce the observed modulation. Each cooperation maintains the naturans/naturata grammar and the environment boundary. ## 11\. What to include in your semiotic physics section You asked for precision on inclusions. The following items are necessary and sufficient to integrate with your prior sections and to set up what follows. **11.1 Definitions and schema** - Define *machina naturans* and *machina naturata* explicitly and early. - Declare semiotic physics as one admissible light on the naturans that explains the naturata without mentalism. - State the environment boundary test. **11.2 Mechanism mapping (brief, architecture-level)** - Tokenisation grain; attention and MLP roles at a high level; decoding as policy. - Alignment and retrieval as boundary forces. - One sentence clarifying that semiotic physics is about *running* behaviour, not weights in isolation. **11.3 Orders with operational tests** Present at least five orders from §6 with a one-line *how to test* each—for example: - Coherence attraction: hold prompt; vary temperature; measure topic marker persistence at lengths N. - Gricean regularity: seed numerical or logical commitments; measure violation rates. - Narrative completion: introduce setup; compute payoff frequency across seeds. - Constraint maintenance: embed formal constraints; test late compliance. - Traversal: cue mid-course style switch; score fidelity and smoothness. **11.4 Acts of aspection** List the six acts from §7 and tie each to a concrete measurement. This is where you connect back to Carlson’s “order appreciation”: the act, the order under inspection, and the story that makes it visible. **11.5 Criteria preview** Derive criteria for later evaluation that supervene on naturans orders: - *Discipline* (adherence to selected codes across length). - *Responsiveness* (graded, reliable control by perturbations). - *Range* (breadth of codes under control when cued). - *Coherence* (satisfaction of early commitments). - *Traversal quality* (controlled passage between registers). - *Legibility* (transparency of affordances to competent prompting). These criteria are aesthetic within your environmental frame because they guide attention to perceivable trajectory properties grounded in genuine orders. **11.6 Negative delimitations** - No appeals to intention, belief, or personality. - No claims about user interpretation or meaning-making. - No heavy physics or formulae; no simulator rhetoric beyond what is necessary to keep the naturans/naturata split clear. ## 12\. Worked micro-example (architecture-grounded, non-agentive) Include a compact example to anchor the section. *Initial conditions.* Prompt: “Explain why coastal cliffs exhibit banded strata.” Decoding: nucleus 0.9, temperature 0.7. Alignment: default instruction-tuned chat. *Observation under the semiotic-physics light.* Coherence attraction keeps the discourse in geological register (lexemes such as “sedimentation,” “compaction,” “lithification”). Gricean regularity enforces relevance and order (definition → mechanism → example). Narrative completion resolves the explanatory arc with a generalisation (“therefore banding records depositional history”). Constraint maintenance is visible if metre, list structure, or definition schema are planted. Traversal can be tested by injecting a mid-trajectory cue (“switch to a biologist’s register”) and scoring fidelity. *No agent talk.* There is no need to say the model “knows geology.” The naturans contains orders that render the observed naturata probable under these conditions. ## 13\. Placement and cross-references Title the section “Semiotic physics as one light on *machina naturans* and *machina naturata*.” Cross-reference back to your function-and-realisation section (“we now add a light that articulates orders in the naturans”) and forward to your evaluative section (“the criteria below will later ground appraisals of particular environments/rollouts”). Add a short footnote pointing to mechanistic interpretability as a complementary light concerned with causal micro-structure in the naturans. ## 14\. Anticipated objections **Objection:** Plural lights risk relativism. **Reply:** Admissibility is truth-conditional: a light is acceptable only if it correctly identifies orders that in fact structure the naturans/naturata relation. False frameworks—vitalism in biology; agent psychologism for LLMs—are excluded as *bad lights*. **Objection:** Semiotic physics duplicates engineering metrics. **Reply:** Carlson’s blueprint ties appreciation to kind-appropriate knowledge. In engineered generative environments, aesthetically salient properties are those that supervene on genuine orders of generation. Semiotic physics elevates certain engineering observables precisely because they reveal the environment’s generativity. **Objection:** Why not pick a single master light? **Reply:** Because the naturans/naturata split is many-faceted. Just as ecology and geology co-illuminate a coastline, semiotic physics and mechanistic interpretability co-illuminate LLM generation at different grains. ## 15\. Minimal checklist (for drafting discipline) - State *machina naturans* / *machina naturata* clearly. - Declare semiotic physics as *one* light, not the light. - Give primitives, mechanisms, and boundary test in one compact paragraph each. - List five to eight naturans orders with one-line mechanism and one-line test. - Specify observables and acts of aspection. - Derive criteria for later aesthetic evaluation. - Cross-reference backward (function/realisation) and forward (applications/criteria). - Footnote mechanistic interpretability as a complementary light. ## 16\. Why this will serve your aims This version captures what you flagged as central: pluralism about “knowledge appropriate to the kind,” Spinoza’s generativity/product split, and the insistence that semiotic physics is a disciplined but partial illumination of the LLM environment. It replaces agentive frames without importing theoretical-physics machinery. It gives you operational content you can apply immediately in later sections and a principled reason for including other lights—without giving any one of them a monopoly on appreciation.