dynamic epistemology

Interview by Richard Marshall.

Sara L. Uckelmanis a hard-core logician and philosopher who thinks dynamic epistemology is the only species of epistemology that matters. When not researching medieval onomastics she thinks about dynamic epistemology alongside medieval philosophy, the obligatio, dubitatio, deceitful agents, the role of formal dialogue systems, Anselm's logic, computability, Llull, Lorhard, ontology and science and why she'd love to be a counter-example to the stereotype of Anglo/American philosophy being dehistoricised. This jive's like taking a time-machine...

3:AM: What made you become a philosopher?

Sara L. Uckelman:This is an interesting question because it presupposes that I am a philosopher! I have railed against that appellation in the past, but now that I've been hired by a philosophy department I feel I must accept the label with good grace. (If you want to know what I am if not a philosopher, I would answer: Logician.) But to answer the underlying question "What made me become interested in philosophy?" there are two answers. One was Plato's Republic, which I picked up, used, in high school, and began to read in public places to project an air of erudition. I didn't understand half of it, but the methodology got me hooked. The second was Lewis's "What the Tortoise Said to Achilles", which I discovered in a philosophy collection from my mom's undergraduate intro philosophy course. I read that, and was captivated. I knew that was what I wanted to do.

3:AM:You're an expert in dynamic epistemology. So what do you mean when
we say epistemology is dynamic and why is it an important species of epistemology? Is this a species of epistemology whereby knowledge of the history of philosophy is really useful in deepening contemporary thinking?

SU:If I'm honest, I'd say it's the only species of epistemology that's important! In practical applications, we are simply not interested in static knowledge, we are interested in how we gain (or lose) knowledge,or how to get other people to gain it, or to hide that we have it, or to pretend that we have knowledge that we don't have. In my view, dynamic epistemology is any epistemology which takes into account any combination of (a) time, (b) action, (c) multi-agent context. It can be dynamic by explaining how a single person's knowledge changes over time; it can be dynamic by focusing on the results of certain actions on a person's knowledge (this will, of course, be closely related to the issue of the progression of time, for I'm happy to be an Aristotelian about time: if there were no actions or events, there would be no passage of time); and it can be dynamic when the creation or transfer of knowledge involves more than one person. (Note that while I speak of 'knowledge', the same goes for 'belief').

Why would we look to history of philosophy, or in my case history of logic in particular, if we're interested in these questions? Because philosophers and logicians in the Middle Ages were interested in exactly the same sorts of issues: What are the different ways that we can know things? What is the relationship between demonstration (proof), testimony, and knowledge? What are the communicative consequences of our knowledge? What is the relationship between knowledge, uncertainty, and ignorance? Entire treatises were written on these topics, and I find them incredibly inspiring to read. The fundamental questions are the same but the ways in which they seek to answer them are different, and provide a fresh way of looking at the issues, and often spark new and innovative solutions.

3:AM:.One area of dynamic epistemology you've interested yourself in is
that of the medieval game-like type of logical disputation, the obligatio. Before looking at this in detail, could you say something about why medieval philosophy was so interested in these games?

SU:I can tell you why scholars of medieval philosophy are so interested in them! There's a famous paper by great historian of medieval logic, Paul Spade, entitled "Why Don't Mediaeval Logicians Ever Tell Us What They’re Doing? Or, What Is This, A Conspiracy?" In this paper, he points out a puzzling phenomenon that "in case after case of the new logical genres that grew up in this later period [13th-14th C], we simply don’t know what was really going on." One such case that he considers are the obligationes disputations. We know what these disputations were like, as we have many 13th- and 14th-century treatises discussing a variety of different types and variations on the rules. But the question of whythey were developed or whatthey were used for has remained perplexingly elusive.

A number of different suggestions have been offered by modern commentators over the years -- myself included -- but none of them have been entirely satisfactory in terms of what the disputations are actually like, and what the treatises themselves say about them. In the past I have supported the view that they are primarily pedagogical in nature, that is, training tools for students. There is good evidence for this view; the treatises themselves say that that they are to provide exercise for students, and we know that these treatises formed a standard part of the undergraduate logical curriculum in the 13th and 14th centuries. Additionally, I have first hand experience with their utility as a training tool, having once had an interactive website where you could play the most basic type, positio, against a computer. (Sadly, the website is no longer maintained).

But this view doesn't provide any explanation for why this -- why did this develop as the tool for training undergraduates as opposed to something else. I am currently developing an argument that will connect the origin of obligationes at the beginning of the 13th century with 12th-century steps towards a non-syllogistic concept of logical consequence, one which doesn't get fully differentiated in the logical treaties until the end of the 13th century. But this is still very much nascent, so I probably shouldn't try to give more detail here!

3:AM:One species of obligation you've investigated is that of the
dubitation which involves giving up knowledge. Can you set the scene for us here what kind of scenarios are we talking about?

SL:It'll be easiest to first briefly sketch the basic variant, positio. In positio, there are two players, the Opponent and the Respondent. The Opponent's job is to put forward a series of propositions, which the Respondent must then respond to by either conceding, denying, or doubting, according to certain rules. (The Respondent is obligatedto follow the rules, which is where the name "obligatio" comes from.) In the basic variant, the initial statement put forward by Opponent is generally false, but contingently so, for example: "You are in Rome" put forward to someone who is actually a student at Paris. In positio, the Respondent is obliged to concede this, if it is not inconsistent. (In the variant depositio, the initial statement is usually something that is true, but contingently so, and the Respondent's obligation is to deny it). Dubitatio is formally like positio and depositio, and some modern commentators have said it is a trivial variant of them, but in fact it involves a layer of complication that makes it much more interesting. In dubitatio, the initial statement is generally something which is know(either to be true or to be false), and the Respondent's obligation is to doubt it. That is, she must not deny anything that the initial statement entails, nor concede anything that entails it (for then by the laws of logic she should also deny the initial statement via modus tollens, or concede the initial statement via modus ponens). But given that there are three possible responses to any statement the Opponent makes, by framing the rules negatively, i.e. "must not deny", "must not concede", instead of positively, the Respondent is allowed a choice: If she must not deny phi, she can either doubt it or concede it. In this way, dubitatio is relevantly more complex than either positio or depositio because it is inherently undeterministic.

One might wonder what relevance this type of disputation might have beyond being a medieval curiosity. In previous work I've argued that the underlying phenomena that dubitatio evidences -- the idea of deceit or pretense in pretending not to have knowledge that one in fact does -- is a relatively ordinary phenomena and one that more people studying dynamic epistemic logic and dynamic epistemology should be interested in. How do we reason when we want to conceal that fact that we have certain pieces of knowledge? Dubitatio gives one a guide to how.

3:AM:So what do we learn about deceitful agents from studying these kinds of cases? How do you analyse what's going on?

SL:From the perspective of building models of what's going on in a dubitatio disputation, the first issue that you run in to is that there is no unique way to move from a situation where knowledge exists to one where it doesn't. For example, suppose you know that "All men are mortal". Suppose you want to pretend that you don't. There are many options open to you: You could pretend that all men are immortal. You could pretend that every man except for Socrates is mortal. You could pretend that every man except for Aristotle is mortal. Whichever option you take, you then have to remember the choice you made, and the ramifications of this can quickly become difficult. So on the practical side of thing, this demonstrates why it is that maintaining a lie is so difficult, and that "what a tangled web we weave // When first we practise to deceive!"

3:AM:And do we gain fresh insights into knowledge by looking at deceit from this perspective, aspects that wouldn't have been noted if the dynamical aspect hadn't been foregrounded?

SU:When develop a method for modeling dubitatio, I found it surprising that standard epistemic systems (both dynamic and static) didn't seem to have any way to account for pretense; in these systems either knowledge is 'hard knowledge', which is irrevocable and cannot be lost, or it is 'soft knowledge', which falls short of actual knowledge in some way. There is no room for have hard knowledge but acting as if you don't. In fact; the phenomenon of deceit I mention in (4) is actually one of a family of phenomena of actions "acting as if". I'm currently developing these ideas with a student of mine, Tom Stephen.

3:AM:I guess these obligations are formal dialogue systems. Was medieval logic informed by them?

SU:Strictly speaking, no, since FDSs weren't developed until the 20th century! But taking the idea of an FDSs a bit more loosely, I'd argue the direction actually goes the other way: The development of the 'Formal Dialogue System' of obligationes informed the development of medieval logic, in particular by helping logicians separate out and make precise the concept of "follows [logically] from".

3:AM:You've written about how Anselm's logic was received in the 20th
and 21st century haven't you. Is modern logic too far removed from Anselm and the medievalists for there to be much interest and dialogue, or is there still interest in Anselm's logic?

SU:Modern logicians suddenly became interested in Anselm as a logician in the latter part of the previous decade when they realized that he espoused a modal view of agency which is very similar to that which can be found in, e.g., stit-theory. Reading some of the early accounts of the 'discovery' of Anselm's modal theory by these agency theorists is quite amusing sometimes; some people seem quite shocked that such a modern approach could be found a millenia ago! I think that this can serve as a good cautionary tale: Never be too sure that what you're doing is new, especially if you don't have much idea about the history of your field! This is one of the reasons why I find studying medieval logic so interesting: So many developments that people think are innovative and ground-breaking nowadays actually have parallels in developments centuries ago. In doing such historical work, one is often reminded of such adages that there is nothing new under the sun, or that those who do not know their history are condemned to repeat it!

[Ramon Llull]

3:AM:How do you connect Llull, Leibniz and Boole? Does this interest stem from your work with your investigations into the history of formal systems of philosophical and logical thinking?

SU:That paper was primarily a side project, something I was asked to contribute to a working at Computability in Europe a few years ago. But it was a project that I was quite excited to do! I don't know very much about either Leibniz or Boole, but Llullis a character and a half. Not only did he write the world's first novel in the vernacular, articulate voting methods that were not articulated again until the 17th and 18th centuries, turn missionary to Africa in his 50s after learning Arabic so that he could convert Muslims, he was also deeply interested in mechanising reasoning, in a way which put him centuries ahead of his time. In a world where time-travel existed, I would love to get Ramon Llull and Alan Turing in the same room together. I can't even imagine what sort of brilliant ideas that would spark!

3:AM:Another area you've investigated is ontology and in particular you've looked at how the word 'ontology' came into being with Jacob Lorhard in the Renaissance. So for the non-philosophers here, can you quickly sketch how ontology is understood by most contemporary philosophers?

SU: In fact, I think I should probably avoid answering this question, because
I don't really know how contemporary philosophers think of ontology, other than that I suspect it wasn't quite the same was as Lorhard thought of it in the early 17th century. Ontology is, for the most part, too far onto the metaphysics side for me -- I like to joke with my students that the border between metaphysics and logic is malleable and movable, but for me, metaphysics begins when logic stops being fun. My work with Lorhard was primarily historical and linguistic -- what was the context in which he used the word? What are the consequences of the diagrammatical approach to ontology that he took. I don't know how these issues are taken up by contemporary philosophers.

3:AM:So what was the background informing Lorhard's use of the term? Was ontology always linked to science?

SU:Lorhard was working at a time where there still wasn't a clear boundary between "science" and "philosophy". There was "natural philosophy", which was the study of nature via philosophical means. It's still unclear why exactly he felt the need to coin this new term, or what he was trying to pick out by it that couldn't be done by vocabulary already in existence. That's another question I'd want to have answered if I had that time machine!

3:AM:How does this link to the growing use of diagrams within conceptual
studies? You've looked at Ramean style diagrams and diagrammatical ontology in general. Can you say something about this and explain whether the classical ontologies are different or comparable with modern ones?

SU:Not really, sadly -- I would love to learn more about the more diagrammatical approaches, both in terms of diagrammatic ontology and also in the role that diagrams play in the development of our philosophical concepts and in pedagogical contexts, but alas I only have one lifetime at the moment, and it is already split between my logical/ philosophical research outlined here, and my other life as a researcher in medieval onomastics.

3:AM:It's clear you're doing hard-core philosophy, and probably you'd be characterised as doing it in the analytic Anglo/American tradition. But that is often stereotyped as a dehistoricised philosophical tradition. Are you a counter-example to the stereotype, and what do you think working with old philosophers and their problems brings to the contemporary thinker's table?

SU:I would be delighted to be a counterexample. I continually find new, inspiring, and exciting things while reading medieval texts. I hope that my research is able to share this inspiration and excitement with others. I think many more people are receptive to learning from history than you might think -- certainly my experience has been that my work gets the best reception when I am talking to modern scholars, rather than to medievalists or to historians of logic. Computer scientists and mathematical logicians especially seem to be genuinely fascinated in the idea that people so temporally remote from us were working on such similar topics. If you think about it, it's actually kind of comforting, and indication that what we're working on are the fundamental questions that transcend time and space.

3:AM:And finally, are there five books you could recommend to readers here at 3:AM that would take us further into your philosophical world?

SU:Absolutely. We're lucky that in recent years, more and more translations
of medieval logical texts are being made into English, so one needn't be a scholar of Latin to have access to the texts. While medieval logical texts have a different vocabulary of technical terms, and are working on some different metaphysical assumptions, the fact remains that what they are doing isn't so very different from what we are doing. (In fact, I recently had opportunity to introduce a number of computer scientists and psychologists to William of Sherwood's Introduction to Logic, written about 1250. "It looks so modern!" one person commented after thumbing through it.) So I can do no better than to recommend that one to to the actual texts:

William of Sherwood, Introduction to Logic, trans. Norman Kretzmann,
U of MN Press.
John Buridan, Treatise on Consequences, trans. Stephen Read, Fordham
University Press.
Eleonore Stump & Norman Kretzmann, eds., Cambridge Translations of Medieval Philosophical Texts, volume 1: Logic and Philosophy of
Language, Cambridge University Press.
Peter of Spain, Summaries of Logic, trans. Brian Copenhaver, Terence Parsons, & Calvin Normore, OUP.
Thomas Bradwardine, Insolubilia, trans. Stephen Read, Peeters.

This is but the tip of the iceberg. In addition to all of the academic reasons for studying medieval logic outlined above, there's also a very strong pragmatic one: There are so many sources out there, and comparatively so few people working in the field, that there is a wealth of material that no one has ever analysed in depth before. I can walk over to pretty much any shelf in my office, and pull off a primary source, flip it open to a random section, and be almost guaranteed to have found material that no one has written on. It makes for a pretty much endless supply of research projects, no small thing in today's publish-or-perish world.

ABOUT THE INTERVIEWER
Richard Marshallis still biding his time.

Buy his book hereto keep him biding!