Sebastian, List, Thank you for responding, Sebastian, and especially for your generous suggestion that my arguments here may already have moved somewhat beyond Davide's early reflections.
I think your response helps clarify an important distinction: You question whether LLMs can properly be redescribed as sign-users given everything that semiosis involves beyond language -- embodiment, action, community, and the human effort involved in developing useful concepts. I share that reservation. However, I don't think my argument requires that an LLM itself qualify as a Peircean sign-user in anything approaching the full sense. What I've increasingly been reflecting on is somewhat different: *Can what an LLM does with signs become incorporated into the larger semiosis by which individuals within a community conduct inquiry? *If AI-generated signs can be interpreted, criticized, corrected, rejected, or developed within a community, then they may contribute to inquiry without needing to attribute to the LLM anything like the full semeiotic agency of the human inquirer. So I was especially interested to read that questions of this sort are part of your project's research. I very much hope you'll continue to bring some of that thinking to Peirce-L; indeed, any facet of your project which you might be interested in discussing here. Best, Gary R. On Sat, Aug 8, 2026 at 2:42 PM Sebastian Feil < [email protected]> wrote: > Gary, thank you for bring the list’s attention to our project. > > We are currently working on integrating the microfilm edition with the > Houghton scans and developing an interface for editing and organising the > manuscript transcriptions (Gemini assisted). Automatic concordance matching > between pages is in a sort of early stage and we hope that it may > facilitate dealing with multiple drafts and especially also with inserted > pages in assembling readable versions of the individual ms. > > My role in the project is more that of a domain expert, as the data > science people call it. I’ve been following your conversation on AI and > semiotics for a while and I have been discussing the “Not minds but signs” > paper with Davide (it was written before the project started). I feel that > your arguments here were already further along than Davide’s early > reflections which may strike large LANGUAGE model people as more > fascinating than those who have been habituated to assume a semiotic state > of mind for a long time now. > > I think the confrontation of minds and signs is difficult to maintain from > a Peircean pov. You could argue that a semiotic extension of the concept of > mind is an advancement from minds as merely linguistic, but I find it hard > to see how the entire range of aspects of embodiment, action and community > that go beyond a merely grammatical conception of semiotics, i.e. how > genuine human efforts end up becoming concepts, and concepts are being > understood as storage of human efforts, would fit into redescribing LLMs as > sign users in the fullest sense. > > However, addressing these questions is part of the research trajectory of > our project and I am excited to see what comes of diving into this. > > Best > Sebastian > > > > > > Am 08.08.2026 um 02:04 schrieb Gary Richmond <[email protected]>: > > List, > > Jon Alan Schmidt directed me to a very interesting and, I believe, > important paper by Davide Picca which he came upon when List member, > Sebastian Fei, sent him an email "providing a link to his current research > project--"Peirce Interprets Peirce" ( > https://peirce-team.github.io/website/). One of the papers already > produced by its principal investigator, Davide Picca, is entitled 'Not > Minds, but Signs: Reframing LLMs through Semiotics' ( > https://arxiv.org/pdf/2505.17080)." > > Jon also provided the Abstract of the paper which I've slightly shortened. > > This paper challenges the prevailing tendency to frame Large Language > Models (LLMs) as cognitive systems, arguing instead for a semiotic > perspective that situates these models within the broader dynamics of sign > manipulation and meaning-making. Rather than assuming that LLMs understand > language or simulate human thought, we propose that their primary function > is to recombine, recontextualize, and circulate linguistic forms based on > probabilistic associations. By shifting from a cognitivist to a semiotic > framework, we avoid anthropomorphism and gain a more precise understanding > of how LLMs participate in cultural processes, not by thinking, but by > generating texts that invite interpretation. Through theoretical analysis > and practical examples, the paper demonstrates how LLMs function as > semiotic agents whose outputs can be treated as interpretive acts, open to > contextual negotiation and critical reflection. We explore applications in > literature, philosophy, education, and cultural production, emphasizing how > LLMs can serve as tools for creativity, dialogue, and critical inquiry. The > semiotic paradigm foregrounds the situated, contingent, and socially > embedded nature of meaning, offering a more rigorous and ethically aware > framework for studying and using LLMs. Ultimately, this approach reframes > LLMs as technological participants in an ongoing ecology of signs. They do > not possess minds, but they alter how we read, write, and make meaning, > compelling us to reconsider the foundations of language, interpretation, > and the role of artificial systems in the production of knowledge. > > > Jon commented on the abstract that "This sounds very much like the > position that you have been advocating recently." I agree, with one small > quibble. > > Although Picca draws extensively upon Peirce, his use of Peircean > semiotics is largely text-centered. While he treats LLM outputs primarily > as signs that elicit interpretants in human readers, he doesn't discuss > as such Peirce's broader conception of *semiosis as an ongoing communal > process of inquiry* in which interpretants generate further > interpretants, and that* habits of inquiry evolve within communities of > investigators*, something which, following Douglas Engelbart's *Intelligence > Augmentation* (IA), I tend to emphasize. But this is a mere quibble. > > This paper largely complements the line of inquiry I'd been pursuing. > While Picca argues, and I agree, that LLMs should be understood as > participants in an 'ecology of signs' rather than as artificial minds, I've > focused on semiosis not being that of isolated agents (human or machine), > but in the *continuing semiotic activity of communities of inquiry *(large > or small). On that view, LLMs may participate derivatively in > intelligence augmentation precisely because *AIs can generate signs that > can become subject to communal criticism, correction, and possible > development*. Picca's semiotic framework points in this direction > although it does not explicitly make that, to my mind, essential Peircean > move. > > Nonetheless, for those interested in the topic, I highly recommend Pica's > paper. > > Best, > > Gary R > > On Fri, Jul 24, 2026 at 9:17 PM Gary Richmond <[email protected]> > wrote: > >> Jon, Vincent, Gary F, List, >> >> Your Subject in introducing this thread, *The End(s) of Inquiry*, has >> rather haunted me these past few weeks since “end” may name (a) the >> * particular* aims of specific inquiries, (b) the *ultimate ideal* toward >> which inquiry tends, or even, (c)* inquiry ending* as we know it. Of >> course, I agree that inquiry cannot be justified merely by its >> continuation, that on Peircean principles that it need be directed >> toward a most admirable ideal, the* summum bonum* of science being *the >> reasonable in itself*. You seem to equate that ideal with truth, >> knowledge, and the growth of concrete reasonableness. No argument there -- >> although I might broaden 'knowledge' to include the growth of >> interpersonal understanding through the development of those habits >> essential to collaborative inquiry. >> >> This is how I might distinguish the *ends of inquiry* from the *end of >> inquiry*. *Particular inquiries* seek adequate explanations of facts, >> valid judgments of arguments, etc. The more *general end of inquiry,* again >> from a Peircean perspective, is the asymptotic growth of understanding >> towards Truth. Truth is *not 'constructed' by us,* as some influential >> French scholars of the mid-20th century would have it, but is something to >> which *we are answerable as a community,* a point which was recently >> emphasized by Gary F and Vincent. On the way, each particular inquiry would >> seem likely to benefit greatly by researchers practicing personal humility >> and discipline, fidelity to the inquiry (and to other inquirers who may be >> involved in the inquiry), and to the hope that *this* inquiry could >> contribute to the growth of knowledge. >> >> As for AI, I agree that responsive systems do not -- and in my view, will >> never -- undergo the 'irritation of doubt' as we human inquirers do. >> Indeed, they possess *no* autonomous concern for truth, or for rational >> conduct, or for moral responsibility. Nevertheless, as I've been arguing, >> *not >> every sign contributing to inquiry must independently confront its >> dynamical object*, that inquiry has always depended on books, diagrams, >> instruments as well as on arguments connecting investigations with >> realities that individual researchers might not have directly experienced. >> Your emphasis on the 'outward clash' is exactly right: The decisive >> question is *whether the larger inquiry remains answerable to Reality, >> typically experiential reality*. Nonetheless, it seems to me that >> AI-generated signs *can* function as potentially valuable interpretants >> after they've been tested, corrected, rejected, etc. Although their Jon, >> Vincent, Gary F, List, >> >> Your Subject in introducing this thread, *The End(s) of Inquiry*, has >> rather haunted me these past few weeks since “end” may name (a) the >> * particular* aims of specific inquiries, (b) the *ultimate ideal* toward >> which inquiry tends, or even, (c)* inquiry ending* as we know it. Of >> course, I agree ythat inquiry cannot be justified merely by its >> continuation, that on Peircean principles that it needs to be directed >> toward a most admirable ideal, the* summum bonum* of science being *the >> reasonable in itself*. You seem to equate that ideal with truth, >> knowledge, and the growth of concrete reasonableness. No argument there -- >> although I might broaden 'knowledge' to include the growth of >> interpersonal understanding through the development of those habits >> essential to collaborative inquiry. >> >> This how I might distinguish the *ends of inquiry* from the *end of >> inquiry*. *Particular inquiries* seek adequate explanations of facts, >> valid judgments of arguments, etc. The more *general end of inquiry,* again >> from a Peircean perspective, is the asympotic growth of understanding >> towards Truth. Truth is *not 'constructed' by us,* as some influential >> French scholars of the mid-20th century would have it, but is something to >> which *we are answerable as a community,* a point which was recently >> emphasized by Gary F and Vincent. On the way, each particular inquiry would >> seem likely to benefit greatly by researchers practicing personal humility >> and discipline, fidelity to the inquiry (and to other inquirers who may be >> involved in the inquiry), and to the hope that *this* inquiry might >> contribute to the growth of knowledge. >> >> As for AI, I agree that responsive systems do not -- and my view, will >> never -- undergo the 'irritation of doubt' as we human inquirers do. >> Indeed, they possess *no* autonomous concern for truth, rational >> conduct, or moral responsibility. Nevertheless, as I've been arguing, *not >> every sign contributing to inquiry must independently confront its >> dynamical object*, that inquiry has always depended on books, diagrams, >> instruments as well as on arguments connecting investigations with >> realities that individual researchers might not have directly experienced. >> Your emphasis on the 'outward clash' is exactly right: The decisive >> question is *whether the larger inquiry remains answerable to Reality, >> typically experiential reality*. Nonetheless, I have argued that >> AI-generated signs *can* function as potentially valuable interpretants >> after they've been tested, corrected, rejected, etc. Although their >> participation in inquiry is clearly derivative and dependent, I have argued >> that it doesn't necessarily follow that their participation in producing >> interesting and potentially interesting interpretants is necessarily unreal. >> >> The deepest danger of AI seems to be for some that if machine generated >> analysis and discourse were to completely replace observation, collateral >> experience, some more-than-human humanistic criticism,etc., then, yes, AI >> could well contribute to the end of inquiry as we know it. However that >> scenario seems *most* unlikely to me. If, on the other hand, responsive >> AI were capable of participating in inquiring by, for example, revealing >> inconsistencies or reorganizing material for testing, etc., I believe that >> it has the potential to benefit inquiry. >> >> More specifically, I have argued that *AI can participate in >> the semiosis involved in inquiry* -- albeit in a derivative, dependent, >> and incomplete way. In this matter I have adopted Doug Engelbart's >> conception of 'Intelligence Augmentation' (AI) as the optimal way to look >> at these new, powerful technologies. Perhaps this goes hand in hand with my >> opposition to any Great Man Theory of Scientific Advancement. (However, >> this is not at all to suggest that there aren't now and haven't always been >> brilliantly creative men and women in all the arts and sciences, people >> whose contribution have been inestimable in moving us towards that *summum >> bonum*.) >> >> I suppose that in a word I would conclude -- and, I believe, on Peircean >> principles -- that it is not only men and women who are capable of >> participating in the semiosis of inqury. AI is a candidate for that >> participation. Quoting Peirce, Jon wrote: >> >> "In logic, it will be observed that knowledge is reasonableness; and the >> ideal >> of reasoning will be to follow such methods as must develop knowledge the >> most speedily." >> >> I think that it is possible that AI can meaningfully help "develop >> knowledge the most speedily." Meanwhile, let's hope that it, and we, can >> contribute to resolving AI's high electricity, water, and space demands >> issues as well as those related to inequalities of access. I'm beginning to >> think that the problem with AI may prove to be much less with the >> technology and much more with the individuals and corporations that are >> developing and controlling it. >> >> Best, >> >> Gary R >> >> >> >> On Wed, Jul 15, 2026 at 5:55 PM Jon Alan Schmidt <[email protected]> >> wrote: >> >>> Gary R., List: >>> >>> The question in bold below is the subject of a new paper that Vincent >>> Colapietro just published online >>> <https://revistas.pucsp.br/index.php/cognitiofilosofia/article/view/76328/50654> >>> in *Cognitio: Revista de Filosofia*, entitled "C. S. Peirce’s Radically >>> Heuristic Conception of Inquiry: Futural Endeavors, Substantive >>> Discoveries, and Sublime Play." As he explains in the text itself, "What I >>> mean by a radically *heuristic *account of inquiry is one in which >>> inquiry is undertaken *for the sake of inquiry*, and this implies that >>> it is undertaken for the sake of the *growth* of inquiry and, by >>> implication, the growth of understanding." Here is the full abstract. >>> >>> In a late manuscript (MS 283 [1906]), Peirce addressed the question, >>> "What are signs for, anyhow?" and raised the possibility that they have no >>> purpose other than being instruments of a *Spieltrieb *(a >>> play-impulse). Peirce however rejected this view and proposed a rival one: >>> signs ultimately serve habituation, the processes by which habits, >>> especially ones more finely and fully attuned to the circumstances of our >>> lives, are formed. But the exercise of our most reliable habits generates >>> sequences of signs apparently displaying a life of their own (e.g., witness >>> the history of science or indeed that of culture). That is, the operation >>> of habits itself seems to be in the service of the generation of novel >>> signs, the furtherance of open-ended processes, destined to transform even >>> some of our most settled habits. Closely connected to this, Peirce appears >>> to slight the monumental achievements of experimental inquiry, treating >>> them as merely the ashes of what was once living science; in effect, he >>> denies their contemporary salience to experimental inquiry. For these and >>> other reasons, the author explores the extent to which inquiry for >>> inquiry’s sake might be seen as an instance of pure play, having its >>> purpose solely in its own perpetuation. While this radically heuristic >>> conception of pure inquiry needs to be qualified, an appreciation of it >>> might require a revision of how the relationship between habits and >>> semiosis is to be understood (a revision in which neither is subordinated >>> to the other). >>> >>> >>> However, it seems to me that there is an important difference between >>> inquiry for *its own* sake and inquiry for the sake of increased >>> *understanding*. Peirce himself identified the growth of *concrete >>> reasonableness* as the *summum bonum*, the ideal end toward which the >>> entire universe is tending asymptotically (though not monotonically). >>> Accordingly, I suggest that the proper end of inquiry is increased >>> *knowledge*, i.e., adopting true beliefs--affirming only those >>> propositions whose dynamical objects are realities, such that the >>> corresponding habits of conduct would not be confounded by any possible >>> future experience. "The purpose of every sign is to express 'fact,' and >>> by being joined with other signs, to approach as nearly as possible to >>> determining an interpretant which would be the *perfect Truth*, the >>> absolute Truth ... [which] is the ultimate interpretant of every sign" >>> (NEM 4:239-40, EP 2:305, 1901). As I have said before, this is what >>> makes logic as semeiotic a *normative *science--it identifies the >>> principles to which any sincere inquiry *ought *to conform. "In logic, >>> it will be observed that knowledge is reasonableness; and the ideal of >>> reasoning will be to follow such methods as must develop knowledge the >>> most speedily" (CP 1.615, EP 2:255, 1903). >>> >>> With that in mind, it seems noteworthy that the dynamical objects of all >>> the signs that "AI" systems generate are *other signs*. Tracing the >>> entire chain of those signs back far enough would *eventually *reach >>> dynamical objects that *humans *actually experienced and then >>> represented in their writings, thereby providing the texts on which the >>> "AI" systems were "trained." However, as Ulysses has pointed out, the >>> latter are always and only processing those symbolic signs *iconically*; >>> and as a doubly degenerate sign, an icon "gives no assurance that any >>> such object as it represents really exists" (NEM 4:242-3, EP 2:307, >>> 1901). Consequently, as I have said before, "AI" systems cannot *verify >>> *the truth of any given sentence apart from probabilistically comparing >>> it with a massive collection of *other *sentences, so it is no wonder >>> that they sometimes suffer from "hallucinations" that produce false or >>> misleading responses to prompts. They lack the "outward clash" that causes >>> the irritation of doubt when our belief-habits *are *confounded by *actual >>> *experience with reality, thereby prompting inquiry. >>> >>> Regards, >>> >>> Jon Alan Schmidt - Olathe, Kansas, USA >>> Structural Engineer, Synechist Philosopher, Lutheran Christian >>> www.LinkedIn.com/in/JonAlanSchmidt >>> <http://www.linkedin.com/in/JonAlanSchmidt> / twitter.com/JonAlanSchmidt >>> >>> On Mon, Jul 13, 2026 at 10:27 PM Gary Richmond <[email protected]> >>> wrote: >>> >>>> Gary F., List,. >>>> >>>> Your examples of the surgical robot and the climate-monitoring AI are >>>> especially helpful in that they help make concrete what we've been >>>> discussing somewhat abstractly. From a Peircean point of view, a crucial >>>> consideration is *that inquiry remains answerable to reality*. Whether >>>> the relevant observations come through human perception, scientific >>>> instruments, robotic and other sensors, or some future combination of these >>>> (and who knows what else?)* reality continues to constrain inquiry* by >>>> resisting inadequate interpretants and interpretations and supporting >>>> better ones. Human collateral experience no doubt remains >>>> paradigmatic, but it is conceivable that future AI systems might >>>> participate more directly in the larger web of observations through which >>>> communities of inquiry remain connected to the world. So, perhaps we differ >>>> a bit here (that is, *should* humanity, let alone life-valuing inquiry >>>> survive the onslaught of planet degrading activity by the 'masters of >>>> destruction' -- you know whom I mean). >>>> >>>> Which brings us to your related question of *alignment*. For even if >>>> future AI systems were to become increasingly integrated into inquiry >>>> through richer forms of interaction with reality, the normative question >>>> would still remain: *To what ends is inquiry directed?* That returns >>>> us to Peirce's architectonic where logic ultimately depends upon ethics and >>>> ethics upon esthetics. (There's much more to be said here, but I happen to >>>> be on vacation in the Finger Lakes district of New York State.) >>>> >>>> Suffice it to say that I too radically distrust "the small group of >>>> owners and developers of AI [who] have purposes which are clearly *out* of >>>> alignment with the purposes and the well-being of life on earth *including >>>> the rest of humanity*." And I too am frankly horrified by AI's "energy- >>>> and water-sucking data centers." As for the first, the "filthy rich" >>>> owners of AI: they already own so much and control so much of what has >>>> become destructive of humanity and the planet that AI would appear to be >>>> just another 'add on' to their greedy lust for wealth and power *if *it >>>> were >>>> not for the second, the already horrific waste of energy and water of their >>>> data centers. While I personally think AI is 'here to stay', I personally >>>> fear for humanity's and the planet's future. >>>> >>>> Best, >>>> >>>> Gary R >>>> >>> _ _ _ _ _ _ _ _ _ _ >>> ► PEIRCE-L subscribers: Click on "Reply List" or "Reply All" to REPLY ON >>> PEIRCE-L to this message. PEIRCE-L posts should go to >>> [email protected] . >>> ► <a href="mailto:[email protected]";>UNSUBSCRIBE FROM >>> PEIRCE-L</a> . But, if your subscribed email account is not your default >>> email account, then go to >>> https://list.iu.edu/sympa/signoff/peirce-l . >>> ► PEIRCE-L is owned by THE PEIRCE GROUP; moderated by Gary Richmond; >>> and co-managed by him and Ben Udell. >> >> >
_ _ _ _ _ _ _ _ _ _ ► PEIRCE-L subscribers: Click on "Reply List" or "Reply All" to REPLY ON PEIRCE-L to this message. PEIRCE-L posts should go to [email protected] . ► <a href="mailto:[email protected]";>UNSUBSCRIBE FROM PEIRCE-L</a> . But, if your subscribed email account is not your default email account, then go to https://list.iu.edu/sympa/signoff/peirce-l . ► PEIRCE-L is owned by THE PEIRCE GROUP; moderated by Gary Richmond; and co-managed by him and Ben Udell.
