Why We Should Be Reading Paul Churchland Right Now

Featured

Posted on by
1

Neurophilosophy and A.I.

The more I get into philosophical and philosophy-adjacent discussions of current-generation “artificial intelligence” (large language models and the like), the more dismayed I am not to see any discussion of the large body of relevant work by Paul Churchland.1 Paul is not ordinarily thought of as a philosopher of A.I., but rather a philosopher of mind and of neuroscience. However, for reasons I hope to make clear in this post, Paul was one of the first philosophers to engage in detail with the predecessors of the technology behind systems like ChatGPT, and he provides quite extensive conceptual resources for beginning to address many of the ontological and epistemological questions about this type of A.I. His work can even help us think about the design of “ethical A.I.” (though perhaps not more practical ethical questions about the use of A.I.).

Paul Churchland is a naturalistic philosopher who has written widely on philosophy of science, philosophy of mind, epistemology, and philosophy of language. For most of his career, the primary naturalistic lens through which he has pursued a variety of philosophical concerns is that of the neurosciences, and from roughly 1986 on, that primarily meant engaging with, interpreting, and applying insights from the emerging science of artificial neural networks (connectionism or parallel distributed processing).2 Beginning in the mid-1980’s, Paul engaged carefully with pioneering work in artificial neural networks by, e.g., Dana Ballard, Garrison Cottrell, Geoffrey Hinton, James McClelland, Andras Pellionisz, David Rumelhart, Terrence Sejnowski, and Ronald J. Williams. Through the next three decades, Paul developed a distinctive interpretation of the cognitive activities of artificial neural networks (ANNs), understood as models of the natural neural networks (NNNs) that make up our brains, and applied that interpretation to issues in epistemology, the metaphysics of mind, the philosophy of science, and moral psychology.

Figure from Matter and Consciousness (2013)

ANNs are constructed by layers of “neurons” that can have a level of activation, which might be represented with a decimal number from 0 to 1. In a densely connected network,3 each number at one layer sends its activation value to the next layer, multiplied by a synaptic weight, which might also vary from 0 to 1, or if inhibitory connections are permitted, from -1 to 1. In a simple “feed-forward” network, the activation value of a node at one layer will be a function of the activation values at the previous layer and the connection weights. In a “recurrent” neural network, some of the connections move in the opposite direction. The first layer will represent some form of input (with each node encoding, e.g., parts of the values of a bitmap of an image, or of the waveform of a sound) and the final layer will represent some form of output (with each node encoding, e.g., part of a classification symbol, or of a motor response).

Networks that start with randomly determined weights will be trained on large sets of data using a training algorithm (e.g., back-propagation, Hebbian learning) that result in a reliable pattern of responses determined by the training data set and the particular training algorithm. The activation of a given layer of the network composed of a number of neurons N can be represented as an N-dimensional vector; each neuron is an axis of the vector space. The transformations from one layer to the next can be represented as a matrix whose dimensions are given by the number of vectors in the two layers. The operation of the network can be computed by a traditional digital computer doing basic linear algebra, the math being a little more complicated for a recurrent network.

Figure from Plato’s Camera (2012)

On Paul’s view, articulated in its most up-to-date form in Plato’s Camera (2012), the best way to represent what the network is doing is by thinking of it in terms of the hypergeometry of these high-dimensional vector spaces. What the synaptic weights that the network learns to represent are the relationships of similarity and difference in the input space as transformed by the preferred or expected output. The way to think of these representations is in terms of maps; these are not two-dimensional maps of a spatial territory, but rather N-dimensional maps of the abstract relations of similarity and difference in whatever phenomena the network is sensitized to. This high-dimensional map is a conceptual space. Concepts are dense regions (attractors) in this sculpted activation space. Cognition is the process of vector transformation accomplished by the connected maps of the network. Ampliative inference is the capacity of such networks to produce an accurate output even when the input is noisy or incomplete, a process known as “vector completion.” Purely feed-forward networks are quite adequate for conceptual spaces for static features, while recurrent networks are more apt for dealing with temporal (or causal) processes.

Figure from Plato’s Camera (2012)

So, for example, an ANN that learns to recognize faces and classify them as male or female comes to represent the abstract space of possible human faces, with distinct regions that represent male and female faces, on a spectrum that includes hyperbolic male and female faces at the extremes, prototypically male and female faces as major attractors, and gender-ambiguous faces at the border between regions. Insofar as the faces in the training data constitute a representative sample of the population the network will later be confronted with, and insofar as male/female represents a genuine dimension of similarity-difference in the population of human faces, we can expect that this map comes to very accurately represent an aspect of the space of possible human faces (in a population) very well.4

From this starting point, Churchland provides us with quite rich accounts of cognition, learning, representation, meaning, knowledge, categorization, universals, explanation, understanding, procedural knowledge, abduction, analogical reasoning, accuracy of representation, the theory-ladenness or conceptual structuring of perception, scientific theories, theory-change, intertheoretic reduction, underdetermination, the aims of science, scientific realism, the cognitive role of (sociocultural) language, situated and extended cognition, metaethics, moral psychology, and more besides. There is plenty of room for debate about the adequacy of each of Paul’s views on these topics, but no room to doubt that in the first instance, the core ideas of the view are worked out for ANNs (for Paul, on the hypothesis that these provide our best current model for what is going on in the mind-brain of actual humans). It is worth noting, then, that whether or not you think ANNs are a good model for the human brain, the philosophical framework that Paul has articulated does apply nicely to ANNs.5 For anyone interested in how we might think about large language models (LLMs) and other contemporary ANN-based A.I., Paul Churchland’s work seems to me to represent a mandatory starting-point.

There is, of course, an obvious objection: the artificial neural networks of 2026 are not the same technology as the ANNs of 1986. The development of deep learning in 2012, generative adversarial networks in 2014, diffusion models in 2015, and the transformer architecture in 2017 were all steps that made the current generation of highly capable “generative A.I.” (GAI) possible. Recurrent networks are mostly out, replaced by transformers and convolutional networks for dealing with temporal tasks. More sophisticated forms of unsupervised and self-supervised learning are used in addition to back-propagation and Hebbian learning for many applications. All of these developments post-date Paul’s major works on the topic. How can we expect, then, to gain fresh insight from Churchland’s work that is now at least a decade and a half old?

Actually, while these developments represent quantum leaps in the technical capability of ANNs, they are mostly insignificant from a philosophical point of view. The fundamental representational capacity of LLMs and other contemporary GAI are still captured by multi-layer populations of neuronal nodes connected by synaptic weights. Generative models and transformers do some of the same work that Paul wanted to do with recurrent networks (in terms of dynamic, nonlinear, and generative processes), which makes GAI less biologically realistic, but for the purposes of Paul’s account, these differences do not seem to me significant. Hebbian learning is still at the basis of many methods of unsupervised learning. Surely, some fine details of Paul’s account will have to be modified, but if one is inclined to think that Churchland’s framework for thinking about intelligence in ANNs is adequate to the systems from the late twentieth century, then it is reasonable to think that they will apply quite well, mutatis mutandis, to LLMs and GAI broadly.

Another relevant difference worth considering is that the kinds of networks Paul typically discusses in his work have perceptual or quasi-perceptual inputs (images, sounds, colors) and perceptual, motor, or linguistic outputs. LLMs have linguistic inputs and linguistic outputs (though multimodal models are becoming increasingly central to the technology). Given his focus on ANNs as a model for brains, when Paul considers language as a part of the system, he is typically thinking about it alongside perceptual and motor modalities. But insofar as corpora of human language have structures (patterns), we should expect that LLMs will embed or comprise cognitive maps of those structures, maps that conceptualize and categorize, know and understand. Suitably capable LLMs can make inferences on the basis of those structures. And if large corpora of human language embed human knowledge about the world, we should expect that the cognitive maps of LLMs will also know something about the world. But insofar as this knowledge is had secondhand, through a medium Paul frequently critiqued for its representational infelicities, Paul also gives us a strong philosophical basis from which to critique, and potentially improve, GAI.

It is very common to see confident assertions that LLMs mimic language use but do not really understand or use it the way that we do, that LLMs do not really reason or think, that they cannot know or understand things. On examination, these claims are often grounded in a folk-psychological understanding about how we think, know, or use language, or, at best, in ideas from philosophy or cognitive psychology that are profoundly disengaged from any understanding of the underlying mechanisms of the brain. LLMs do not reason, the thinking goes, because reasoning is an ordered, linguaformal process that conforms to logical or epistemic norms; they do not know because justification must be based on underlying world-models which LLMs lack. But if Churchland is right, reasoning just is the transformation of vectors in a high-dimensional activation space that maps the relevant cognitive realm; that activation space just is the relevant world-model (map) that the LLM relies on in understanding the world.

So Paul might have it, I think. Maybe these views won’t hold up to closer scrutiny, but I hope it is clear why his is a starting point worth engaging with, if we hope to make sense of the idea that artificial neural networks like ChatGPT are “intelligent.”

Notes

  1. In the spirit of full disclosure, Paul was my dissertation chair. I will mostly refer to him as “Paul” throughout this piece, both because of familiarity and to distinguish his contributions to this issue from those of his spouse and frequent collaborator, Patricia Churchland. ↩︎
  2. This marks an important division of labor between the work of Paul and Patricia Churchland. Pat’s work has tended to focus more on neurobiology and Paul’s more on neural network models, a field at the intersection of computational neuroscience and computer science. Even Patricia Churchland’s and Terrence Sejnowski’s The Computational Brain is much more biologically-grounded than most of the work that Paul relies on and interprets. The division of labor is of course, porous, given the frequency of their collaboration and the degree to which they influence each other, but it is nonetheless a clear difference to those familiar with both of their ouvres. ↩︎
  3. In a sparsely connected network, the neurons at layer N will only send signals to some, but not all of the neurons at layer N+1. ↩︎
  4. Whether and to what extent this particular example, or the view at large, provides resources for, or a target for, feminist critique are left as an exercise for the reader. ↩︎
  5. Pace some small quibbles concerning Paul’s account of connectionist networks. See Lasko and Cottrell, “Churchland on Connectionism” (2006) in Paul Churchland. ↩︎

Internet History, Media Archaeology

Featured

Posted on by
Reply

Hello, Blog. It has been a while.

Mainly, I’m writing this to share a couple of videos I came across this morning, that I enjoyed, a couple of Vlogbrothers entries from the past few days. The Brothers Green are not everyone’s taste, but I usually enjoy them, and these two videos in particular tickled my fancy.

First, Hank Green, “The Forgotten Viral Video that Could Never be Made Today”

Second, John Green, “You have no chance to survive make your time”

I like these little glimpses into Internet Past, and probably one of the things that I enjoy about Hank’s and John’s work is that (even though I did not start following them until COVID times), they’re Early Internet people, with whom I feel a lot of kinship. And one of the things I like is how these kinds of discussions help us think about not only how we got where we are, but the contingency of the current way of things, which also entails the possibility of a different future than the trajectory we seem to find ourselves on. (Which, when it comes to technology, the internet, and media, seems like a trajectory that mostly sucks.)

I am not sure when exactly I became fascinated with technology/digital/media history, but I can point towards a few things. In 2009 I was hired at UT Dallas, primarily to teach Philosophy of Technology to undergrads in the relatively new Arts and Technology (ATEC) major. Despite what I said in my application letter, I didn’t know much about philosophy of technology, so for the first few years, I taught a pretty standard version of it (Heidegger and Marcuse, oh my). During these first few years, I became increasingly connected with the Emerging Media and Communication (EMAC) group, which created a new degree program out of ATEC.

At some point in late 2012 or early 2013, I encountered Anil Dash‘s “The Web We Lost,” which also has a video version:

I was an early adopter of Movable Type for my blog back in the day, and have enjoyed Anil’s ideas from all they way back. At this point, I was very much starting to worry about how the Open Web was being replaced by walled-garden social media, a trend that has only gotten worse over the decade+ since then.

In 2014, I redesigned Philosophy of Technology to better fit the idea of thinking critically about the internet and digital technologies that I wanted to teach ATEC, EMAC, and the odd engineering and computer science student who ended up in that class. I included a section on “Lessons from the History of Technology” that included Neil Stephenson’s In the Beginning… was the Command Line and Ted Nelson‘s cranky Computers for Cynics YouTube videos.

This led to a session I ran with Andy Famiglietti at the 2015 Digital Frontiers conference, “Silicon in August: In search of a usable past for alternative digital futures” (sad note that the “Digital Frontiers” website exists only in the Internet Archive). This was a session where we asked a lot of questions, but didn’t come up with a lot of answers. It was fun, and interesting, and I thought it might lead somewhere.

One of the reasons that I study the history of philosophy, and one of the things that I think is interesting about integrating the history and philosophy of science, is that it provides us with a sense of perspective, a sense of the contingency of the questions, approaches, and answers that we focus on in the present, and allows us to have a broader view of where we might go in the future. I like how Dewey’s pragmatism breaks us out of the narrow bounds of contemporary professional philosophy, and how Marston’s weird century-old ideas about psychology can help us think differently about what contemporary psychology and cognitive science could be. Hasok Chang’s “complementary science” has been a huge inspiration for me in this respect.

I wanted to bring the same approach to digital technology. Ultimately, I wanted to start working with students to dig into the material histories of these systems, including getting hands-on experience with old computers, old software, old communication systems, very different from the ones they were used to in their day-to-day lives, as a way of starting to think about alternative futures. But things changed, people moved institutions, programs appeared and disappeared, and my work down that road did not really take off.

Early this week, my good friend Bryan sent me a link to the Media Archaeology Lab (MAL) at CU Boulder, run by Lori Emerson (a colleague of our mutual college friend Professor Casey). They have been doing since 2009 something I start thinking about only around 2014, research and teaching the history of media and technology through hands-on material interaction and recreation. Some of the work there is truly fascinating, and I strongly recommend you lose an afternoon looking around MAL’s and Professor Emerson’s websites. (Somewhat similar work was also being done at UT Dallas by Anne Balsamo and Sean Landers before I left, and I wonder if I would have picked this up again had I stayed in Dallas instead of moving to SIU in 2022.)

There never was an attempt at a robust history and philosophy of technology to the same extent that there has been such an attempt with the history and philosophy of science (not that the partnership there has been a successful and happy one). I think in part that it has to do with very different attitudes towards the subject by historians of technology and the philosophers of technology. But I do think that contemporary philosophy of technology and the developers of contemporary technology could do better to understand the history, understand what we have lost, to help us find a better way forward.

Duck Genitals and Feminist Science Studies

Featured

Posted on by
Reply

Spring 2013 saw another round of misguided right-wing attacks on basic scientific research in the U.S. Congress, a political tactic that purports to demonstrate the wastefulness of the federal government by showing off the price tag (often small in terms of scientific research budgets) for obscure research that can be described in ways that make it sound goofy or idiotic. This time around, it peaked my interest a good bit more, because it brought national media attention to one of my favorite bits of biological research: Patricia Brennan’s work on duck genitalia. (Brennan wrote a wonderful defense of her research for Slate. Even the Union of Concerned Scientists weighed in.)

Why do I love this research so much? The biology is interesting, yes (more on that in a minute), but also, as a philosopher of science with a long-standing interest in feminist science studies, I see it as following the exact structure of some of the classic cases from that literature. That is, Brennan’s work exemplifies the pattern of research of women entering a field of research dominated by men, revolutionizing and improving the methods and theories in that field. It is thus similar to the earlier cases of primatology as described by Donna Haraway—where scientists hadn’t paid much attention to the behavior if female primates and ended up with theories where their roles were entirely passive—and reproductive cell biology as described by (inter alia) Emily Martin—where the “Prince Charming/Sleeping Beauty” theory of sperm/egg fertilization was a going idea, I kid you not.

To get the basics, let’s start with this “True Facts” video by Ze Frank:

Continue reading →

Philosophy, Funding, and Conflict of Interest

Featured

Posted on by
9

A couple of weeks back, Justin Weinberg at the Daily Nous posed a really interesting question. The context was Daniel Dennett’s review of Alfred Mele’s book Free: Why Science Hasn’t Disproved Free Will. Dennett gives a relatively standard story about conflict of interest in science funding using a hypothetical story of saturated fat research funded by “the Foundation for the Advancement of Bacon.” On standard accounts, we are right to apply a higher level of scrutiny towards research whose funding displays a potential conflict of interest, and this is why, e.g., we have COI reporting requirements in certain journals and for federally funded research.

Dennett then points out that Mele’s work is funded by the John Templeton Foundation, which (simplifying a bit) has an ultimate agenda the integration of science and religion, and lately has been funding large projects that involve philosophers, scientists, and theologians working together on a shared theme, like Free Will or Character. Mele has received and managed two such grants.

Here’s Justin:

Mele’s project is not the only Templeton-funded philosophy project, nor is Templeton the only source of funds with an agenda. Dennett is claiming that funding from ideological sources casts a shadow on philosophical research in much the same way that funding from industry casts a shadow on scientific research. Is he correct?

Unfortunately, the question was lost as the thread got hijacked by a lot of nonsense, specific details about Templeton and Dennett’s neo-Atheist anti-Templeton agenda, as well as some understandable pragmatic implications of Dennett’s statements on Mele’s character. Most egregious were the many denials that conflict of interest is an issue in science, that they somehow amounted to a fallacious ad hominem argument. For instance, Terrance Tomkow and Kadri Vihvelin claim “the motives of the researcher or his employers are always beside the scientific point.” Dennett answered this point well when he said,

As for Tomkow and Vihvelin’s high-minded insistence that one is obliged “to ignore” the sponsorship of research, I wonder what planet they have been living on recently. Why do they think that researchers have adopted the policy of always declaring the sources of their funding?

Or as Richard Zach said, “It’s as if the past few decades of work on values in science didn’t happen.”

I think Justin’s original question is interesting, though, because it encourages us to think past the specific details of Mele’s book, Dennett’s critique, and the Templeton foundation. Maybe it is because I work at a STEM university, but I often hear talk that the humanities are going to have to move more towards extramural funding. For philosophers, Templeton is where the big money is, but there are also plenty of smaller private foundations, donors funding endowed Chairs (as Zara pointed out), and so on. It’s a timely question. And it is one that invites us to reflect on the similarities and differences between the sciences and philosophy (or the humanities more broadly). I wish more commenters had taken up the call.

I would suggest one analogy and one major disanalogy between science and philosophy in regards to conflict of interest. The analogy is, if I understood him right, what Dennett was getting at: funding applied on a large scale can alter, or even distort, the research agenda of a discipline. And evaluating that will require us to think about what the research agenda ought to look like.

The importance of research agendas in science is the centerpiece of Philip Kitcher’s Science, Truth, and Democracy and Science in a Democratic Society. He describes the ideal research agenda for science or a scientific discipline as well-ordered science (WOS), and he argues persuasively that not only epistemic and internal disciplinary values, but also practical and ethical values are central to determining what counts as WOS. Further, he argues that WOS should be evaluated democratically, in some sense. Because science is a social institution, it is ultimately responsible for serving the public. Kitcher also rightly recognizes the roles of funding sources and individual choices in actually setting research agendas, and argues that individual sciences have a duty to stand up and fight for change when the research agenda in their field is very far from well-ordered.

Likewise, we could ask about what “well-ordered philosophy” would look like. Presumably, many philosophers (like many scientists) would argue that notions of intrinsic intellectual/philosophical merit, strength of argument, and freedom of research should determine the ideal research agenda. I, and I suspect Kitcher as well, would prefer pragmatic, ethical, and political considerations to play a role. Either way, we can ask whether and how funding sources are moving us towards or away from a well-ordered research agenda.

Mele’s work discusses Free Will, argues that contrary to some triumphalist claims, the sciences haven’t settled the question yet, criticizes some of those claims by scientists, and is agnostic about whether free will is compatible with determinism. I’m not sure how those things fit with the ideological agenda of Templeton, though I can understand the feeling that they do, somehow. And insofar as Templeton wants to stay a major player in funding research on Free Will, we could see more of this sort of thing, less of other approaches. Zooming out to the context that Justin invites us to consider, it is worth wondering what the effects of funded research can be on the research agenda of philosophy, and it is worth deliberating about whether some funding sources should be considered a problematic conflict of interest, Templeton included. (My own view, held tentatively, is that Templeton is alright in this respect but should be closely monitored.) But also note, that until one has a sense that funding agencies are having a systematic effect, it doesn’t seem reasonable to criticize individuals in the way that Dennett does (if implicitly).

The disanalogy I would like to mention has to do with the different types of arguments that are made in empirical science and in philosophy. Philosophical arguments are usually scholarly while scientific arguments are generally technical. I mean these in a specific sense inspired by Bruno Latour’s Science in Action (N.B., these terms aren’t the ones Latour uses). To make an argument in philosophy requires nothing more than a library card, writing implements, the ability to adopt the current style of the literature in the field you wish to contribute to, and the ability to craft an argument. Scholarly arguments can be evaluated on their surface—you need only to examine the content of the text itself, and perhaps the cited sources, to understand the argument or produce a counter-argument.

Some elements of scientific texts can be evaluated in this way. But scientific arguments are also technical. In particular, much of the argument hangs on what Latour calls inscriptions—tables, charts, graphs, and figures—produced by instruments. There are hard limits to how far one can interrogate a technical text. One can raise questions about certain inferences and interpretations, and one can examine the equipment and materials that produce the data and the inscriptions, at least, as long as one has an invitation to the relevant laboratory and the patience of one’s host. But past a certain point, making an effective counter-argument requires a counter-laboratory with instruments producing inscriptions that can be used in arguments. To a large extent, the technical nature of modern science is a major source of its power and effectiveness; but a cost is that we have to rely on trust to a greater extent. And conflict of interest is at least a pro tanto reason to withhold that trust, whereas trust is not at issue in philosophical arguments in the same sense.

So while it is incorrect for Jim Griffis to say that “If the ‘science is impeccable, carefully conducted and rigorously argued’ there would be no problem with who paid for the research,” because of the technical nature of science, he is right to say that “for philosophical works, either the argument is cogent or it’s not.”

Full disclosure: I have previously applied for (but not received) Templeton funding.

Wonder Woman’s Lasso of Truth

Featured

Posted on by
1

I’ve just begun reading Jill Lepore’s new book about Wonder Woman and William Moulton Marston. So far, I’m finding it to be really thorough and excellent! [Edit: My final assessment was much more mixed.] I was a little disturbed, though, to discover that on the first page of the preface, Lepore makes a basic mistake about one of the key features of the early Wonder Woman comics:

“She had a magic lasso; anyone she roped had to tell the truth”(xi).

She repeats the point in one of the color plates in the middle of the book, which does show Wonder Woman compelling a thug to tell the truth. The accompanying text that connects this work to lie detectors and Marston’s work on deception is ultimately misleading, however.

What’s the problem here? Isn’t it called “the Lasso of Truth?” The problem, as Brian Cronin pointed out a couple of years ago at Comic Book Resources, is that this is actually anachronistic. Marston called this iconic element of Wonder Woman’s gear the “Magic Lasso” or sometimes “Golden Lasso,” not the lasso of truth. And its power has nothing specific to do with the truth, but rather with compelling obedience.

Lasso1

LieDetectorAIt’s a tempting connection to make Marston, after all, invented the lie detector test, or at least, he’s one of its most recognizable developers and proponents. It’s a common and tempting connection to draw:

Here’s Geoffrey Bunn, one of the few historians of psychology to write in detail about Marston:

“Anyone caught in the lasso found it impossible to lie. And because Wonder Woman used it to extract confessions and compel obedience, the golden lasso was of course nothing less than a lie detector.” (Bunn 1997, p. 108)

slaveThe real story behind Wonder Woman’s magic lasso is much more interesting and much stranger. Marston was an experimental psychologist who developed a theory of emotions. According to his theory, the four basic emotions were Dominance, Compliance, Inducement, and Submission. According to Marston, submission was a matter of giving over one’s will to a basically friendly stimulus; not only was it necessarily a pleasant emotion, but it was a necessary component of love and thus of a healthy psyche. What the magic lasso was able to do, it seems, was to place the person bound in an automatic state of submission to the will of the lasso’s wielder, making them happy to do whatever you asked. Including, occasionally, to tell the truth when they intended to deceive. Lasso2

However, more often than not, when Wonder Woman wanted to know whether someone was telling the truth, she’d make use of the very tool that Marston invented for that purpose, a lie detector test based on systolic blood pressure measurements.

LieDetector1

 

Above I called the mistake an anachronism, because while Marston never used the term “Lasso of Truth,” present day comics do refer to it by that name. According to Cronin, this usage began in Wonder Woman volume 2 #2 (1987; Writer: Greg Potter, Artist: George Pérez, Editor: Karen Berger). This is the post-Crisis reboot of Wonder Woman, meaning that it occurred after the Crisis on Infinite Earths mini-series that altered the continuity of the DC Comics universe. Presumably, the creators knew about Marston’s interests in lie detection, and decided to change the powers and name of the lasso accordingly. (On the other hand, the commenters on Cronin’s piece suggests that the usage comes from the Wonder Woman TV show starring Linda Carter, so perhaps the connection was made by the creators of the show.)

UPDATE: Noah Berlatsky gets it right in his new book, Wonder Woman: Bondage and Feminism in the Marston/Peter Comics, 1941-1948.

Besides her superstrength, superspeed, superendurance, and other physical prowess, [Wonder Woman] also has bracelets that she can use to block bullets, an invisible plane, and a magic lasso that compels obedience to her commands (in later iterations, the lasso’s power is often downgraded so that it forces people to tell the truth rather than forcing them to obey any command).

What is “The Hanged Man”?

Featured

Posted on by
2

“The Hanged Man” was an online Synchronet BBS Home Screen alias or “handle” I adopted somewhere around 1994, when I didn’t even have access to the internet and instead was using local dialup bulletin board systems (BBS’s). I continued to use the name on into the next millennium, when I started a webpage and got an email address. I’ve continued to use it into the present mostly out of inertia. It’s also a fairly memorable handle, and I’ve had the web domain long enough not to want to give it up.

Continue reading →

The Loss of Hope

Posted on by
Reply

This is going to sound a little silly at first, but bear with me.

Growing up in the 1980s and ’90s, I was taught that the world was becoming more “civilized,” that post-WW2, wars would only happen in backwards places, that the world was no longer safe for fascism, and that, with the end of the Cold War, a new millennium of peace was approaching. As a young adult in the 2000s, the narrative was that terrorism was the new threat, and I protested war in far-off places conducted on false pretenses. Even then, those unjust wars were more precise, more targeted, somehow more “civilized.” We were somewhat more disillusioned, but there was still the feeling that the horrors of the prior century were firmly in the past. The 2010s kicked off with the Arab Spring, and our hope-and-change administration won a major victory for domestic policy, and it seemed that technology and democracy were advancing the world over.

It feels like in the last 10 years all the hope and most of the progress of the prior century have evaporated. There is a protracted land war in Europe. Our most “civilized” ally in the Middle East has shown itself capable of bottomless barbarism. The technologists who seemed to be heroes ushering in utopia have proven themselves to be cartoonish villains who have brought us a cyberpunk dystopia – or not even that, because it isn’t even stylish in the way that fictional cyberpunk dystopias were. Neofascists (or something equally bad) have come to power in the US and in many other previously liberal democracies, and they are in the process of dismantling democracy, to replace it with autocracy and oligarchy. They are sending the military into American cities. The party of hope and change has shown itself to be instead the party of timid acquiescence.

I’m describing a sort of structure of feeling and how it has changed over the course of my life. It is, of course, told from my immensely privileged, relatively wealthy, white, American, male perspective. Many people around the world born the same time I was, from poor Black Americans to Afghani villagers to Chinese factory workers could tell you that from the beginning the story of the “civilized” age was a lie, thick with the ideology of capitalism and US supremacy. And yet… and yet, I still feel that something really important has been lost, not just for people like me, but for the entire future.

There is no Analytic/Continental Divide in Philosophy, and there Never Was — A sketch of an argument

Posted on by
Reply

Who was the German philosopher and mathematician writing in the late 19th and early 20th century, who explored the philosophical foundations of mathematics and logic and argued against theories that would reduce mathematics, logic, or meaning to psychology? He posited a theory of meaning dependent on both the reference and a non-psychological, non-representational, objective “sense” or “intentional content.” — If you said Frege or Husserl, you would be correct.

Who was the early-mid 20th century German philosopher who was heavily influenced by neo-Kantianism and Husserl, who wrote early work on logic, highly praised Frege and Nietzsche, and worked towards the overcoming of traditional metaphysics, which he argued had become an obstacle to philosophy? — If you said Carnap or Heidegger, you would be correct.

Which early-mid 20th century philosopher theorized the radical finitude of human beings, the limits of reason, the impossibility of reducing language to a set of rational rules or finding unshakeable philosophical foundations for our thinking? Who argued that philosophy is often confused because it begins from an intellectualized, theoretical position rather our everyday involvement with the world? — If you answered Heidegger, Wittgenstein, or Dewey, you would be right.

Which thinker sharply criticized the obscure jargon of Heidegger and his followers as an abuse of language, and one with problematic political consequences? — Yes, if you said Carnap or Adorno, you are correct.

These overlaps and connections hint at the possibility that the “analytic/continental” divide, which has loomed large in our understanding of 20th century philosophy and into the present, has been exaggerated, or at least that its nature has been misunderstood. Relatedly, Bernard Williams has referred to the analytic/Continental divide as “a strange cross-classification — rather as though one divided cars into front-wheel drives and Japanese.” It raises the question of whether the divide is part of a philosophically respectable categorization scheme, or a coherent narrative, for understanding our discipline.

I will here sketch a brief argument that there is not and has never been a genuine philosophical divide between analytic and Continental philosophy. That is not to say that there is not some kind of sociological or political divide; only that the characterization of that divide as having some serious grounding in disagreements about philosophical methods or views is a mistake, a kind of false consciousness or ideology.

The purported divide is not a dichotomy

Throughout the entire period, there have been practicing philosophers who are neither fish nor fowl: pragmatists, Whiteheadian process philosophers, liberal naturalists, traditionalists, etc. For a distinction that is meant to divide the discipline, there are an awful lot of philosophers who fall between or sit aside from the distinction.

Indeed, I myself do not identify with either. My undergraduate training in philosophy was in a very small program with a traditionalist historian of philosophy, an avowed pragmatist, a couple of philosophers of science with no sympathy for traditonal analytic methods, and a physicist with a fascination with Whitehead and Buddhism. My graduate program certainly had self-identified analytic faculty, but many of the faculty were traditional historians of philosophy or considered themselves post-analytic, post-empiricist naturalists, Marxists, or bridge-builders who explicit embraced ideas across the divide.

Neither side of the divide is a unified thing

Neither analytic nor Continental philosophy has any (philosophically substantive) identity. Neither term was a self-identification at first — “analytic philosophy” was a term first posited by traditionalist and non-analytic American philosophers to describe a foreign phenomenon (some friendly, some hostile). Similarly, British philosophers started referring to “continental philosophy” largely to distinguish themselves from a form of philosophy to which they were hostile. What family resemblances do bring larger aspects of each “side” together are also not sufficient to include and exclude everything appropriately. (Popular reference to different readings of Kant or emphases on the first vs third Critique captures relatively little of either side.)

On the analytic side

Analytic philosophy historically included approaches as disprate as Cambridge realists, logical atomists, logical empiricists, ordinary language philosophers, among others. And all of these former schools are moribund; there is even less uniting the contemporary work of supposedly “analytic” philosophers. “Analytic” has become a term of either abuse or affiliation, depending on who is using it, for most of what is currently popular in mainstream philosophy.

On the Continental side

Continental philosophy supposedly includes German idealists, phenomenologists, existentialists, hermeneuticists, Marxians, critical theorists, speculative realists, and more. (Though not all of them, for some Marxists, phenomenologists, scholars of German idealism, etc. are considered “too analytic.”) The schools have little in common and are at odds with each other as much as with any analytic approach.

Analytic and Continental philosophy, even if they exist, are not divided

The two supposed traditions share the same origins in late 19th and early 20th century debates concerning anti-psychologism and neo-Kantianism. The major founding figures of each tradition were in dialogue and influenced one another, and there have been many exchanges (some fruitful, some less so) over the years.
The sense of unbridgeable divides is typically the result of honest misunderstandings and bad faith misrepresentations in these exchanges, rather than substantive philosophical differences. Where those differences exist, they are not more significant than “internal” disputes within the supposed traditions.

For example, many (not all) analytic and Continental philosophers take some version of the linguistic turn, a point that distinguishes them from most other philosophers but not from each other. Indeed, in this and many other ways, analytic and Continental philosophy are more like each other than they are like other forms of philosophy. Another example would be their anti-naturalist origins, where the founding “analytic” and “Continental” figures took essentially the same positions in the Psychologismusstreit.

The divide is increasingly porous

Today, there is significant traffic across the supposed divide. To name just two examples: (1) the influence of phenomenology in analytic philosophy of mind; (2) the influence of Sellars on continental realism. Such mutual influence has only been for the better. To whatever extent a divide ever existed (sociologically), today it is only a roadblock to better philosophy.

Why does it matter?

The divide obscures our understanding of the history of philosophy

As historians of philosophy like Michael Dummett, Andreas Vrahimis, Abraham Stone, and Carl Sachs have shown, there are significant interactions, influences, connections, and encounters between figures across the analytic/continental divide. Setting up the divide, and treating as if it is something that has been with us for centuries (since Kant or even before, as some would have it), only serves to obscure those connections. Accurate reading in the history of philosophy requires awareness of context. If you are reading Husserl and Heidegger, but not Frege and Carnap, you don’t really understand what Husserl and Heidegger are up to.

What’s more, the sociological divide at its earliest is a mid-twentieth century phenomenon, but it is often read backwards to inflect readings of earlier thinkers. Our understanding of figures like Kant, Hegel, and Marx becomes distorted by anachronistically reading contemporary concerns and divisions into their work.

Partisanship to the divide narrows who we read and who we teach in a harmful way

The divide has created separate canons of philosophy, who is read and taken seriously. The more our education hews to the divide, the more blinkered the canon becomes. This is especially harmful when we think about figures (Cassirer, Dewey, Whitehead) who are outside the divide.

The divide hinders progress in philosophy

Some of the truly groundbreaking and impactful work in philosophy has bridged the divide, drawing on both “sides”. This work is harder so long as students are discouraged from reading across the divide and work that engages across the divide meets with hostile boundary police.

So why did we believe in the divide?

If there is not and never was a genuine philosophical divide, why did we think there was? Any argument of this sort owes an explanation. Here are a few possibilities:

  • The perceived relation between German national identity and German philosophy in light of the World Wars, due to flat-footed Anglophone critiques.
  • The impact of Nazism on Jewish and socialist philosophers in Continental Europe, many of whom resettled in the Anglophone world.
  • Differing perceptions of Heidegger’s Nazism and its separability from his philosophical insights.
  • The professionalization of philosophy and different reactions to it by different groups of philosophers.
  • The Cold War & McCarthyism undoubtedly played some role, marginalizing certain traditions and transforming others (though sometimes this role has been overplayed or its significance misunderstood).
  • The reception of relatively marginal French philosophers in prominent British/American literary theory circles.
  • Misunderstandings as a result of philosophers of different languages or writing in different styles talking past one another.
  • Institutional trauma working post hoc to reify a sociological divide by rationalizing it into a significant philosophical divide.

These are sketch answers, but worth considering further.

September 2020

Posted on by
Reply

It is a week out from my birthday, and I’m feeling a bit reflective. It has been quite the year. For me, a year of personal, professional, and numerological milestones and achievements. For the world, a year of tragedy, fear, social isolation, and a host of adjustments to new ways of working, teaching, learning, and living.

At the end of March I received the news that, effective September 1, I would be promoted from Associate Professor to Professor. So, for the last few days, I have officially been a full Professor, the highest “normal” rank for a faculty career.1 In a very real sense, this is all I have ever wanted to achieve professionally for the past 20 years or so, when I started to get the idea that college professor was a job one could have, focused mainly on thinking, reading, writing, and teaching. And I did it at 39 years old.

This year is my 10th as director of the Center for Values in Medicine, Science, and Technology. It was an exciting if terrifying duty to take on as a pre-tenure Assistant Professor, with a steep learning curve. I think in 10 years I have been able to use this opportunity to make some small, positive impact on the field of philosophy of science (and some adjacent fields), on some members of the local community, and on the education of our students. It literally occupied my entire 30s. There are still always new challenges.

Originally, I had hoped my book would come out before my birthday, but some understandable delays make it look more like late October. Still, my first book is an exciting accomplishment from my perspective, and not one that is just a matter of course for someone in my field (plenty of philosophers of science write only articles). I hope it will have an impact, and I am glad that I was able to find a source of funding that would make it Open Access.

In addition, I have for the last year (and will for the next two) been Program Head / Program Coordinator for History and Philosophy (sort of Department Chair Lite). I feel like I’ve done some good work here as well, helping reform our graduate curriculum, winning greater autonomy for our programs from a previously very top-down system of governance, improving the sense of faculty ownership of the program, replacing our just-in-time scheduling model with one of (more but not perfectly) fair and balanced rotations of duties and better choices for students. I’ve helped build a Philosophy B.A. program from scratch and expand the HIST and PHIL offerings in the gen-ed core curriculum. That said, there have been growing pains, and some disorganization and mistakes on my part, but I am lucky to work for a really great faculty who have stepped up to work hard on this stuff. I’ve also been lucky to follow the good work of the person who had this job before me and to have great people as heads of other programs in the School to work with. The pandemic situation has added a lot to the job that I did not expect. I will be relieved to pass the torch when my turn is done.

Of course, most of the major achievements of this year have been somewhat anticlimactic. I wasn’t really able to celebrate my promotion, and I won’t be able to have much of a party for my 40th next week. Many of the Center’s and the History/Philosophy Program’s major events were cancelled, and it is not clear when we’ll be able to have a live lecture or a conference again. Teaching is both harder and less satisfying than ever. And so for all that it has been a big year, it has also been a melancholy one. (That said, though I’ve taken some knocks from 2020, I know I’ve also been very lucky.)

The big questions are: where to go from here? Having achieved everything I had wanted professionally by 40, where will I set my sights next? That’s something that I’ll definitely be thinking over in this month of reflection.

  1. One can from here win one of a limited supply of named or distinguished professorship or chair positions, or move into the administrative track, but both moves are relatively rare.↩︎

Wishful Thinking and Moral Imagination

Posted on by
Reply

A bit of writing that I was doing for my book went on a tangent that I really cannot incorporate into the book. But I decided to follow it anyway, because it was fun and interesting and kind of weird, to see where it lead. Well, it led me into almost 2000 words of writing, below.

There is no life I know
To compare with pure imagination
Living there, you’ll be free
If you truly wish to be1

In a variety of contexts, “wishful thinking” is used as a pejorative. In politics, wishful thinking is a charge political realists levy against radical reformists who want more than gradual and piecemeal change, who do not subject their calls for reform to cost/benefit analysis and the like. In another area, those who defend an apolitical, value-free conception of science argue that feminist philosophies of science and other value-laden approaches give license to replace scientific evidence with wishful thinking. Much work in the literature on values in science lately has been devoted to refuting this claim, arguing that value-laden science need not engage in wishful thinking, that it remains objective (Longino, Harding, Douglas), that introducing values makes science more demanding, not less (Kourany). Of course, some values influencing some science in some ways will amount to wishful thinking, but these account seek to rule such science out of bounds, epistemically or even ethically.

What if we, instead, question the universality of the negative evaluation of wishful thinking. Isn’t wishful thinking, after all, connected with our capacity for imagination, which is important to the creativity that drives scientific discovery, as well as to the moral imagination that, I would argue, is central to our ethical decision-making and our moral life? I think so, and if so, there are reasons to mount a moderate defense of wishful thinking in some sense.2 Indeed, we can look to a wide variety of artistic works that hold up the power of wishful thinking.

Continue reading →

Notes on Anthony Weston’s How to Re-Imagine the World

Posted on by
Reply

Anthony Weston‘s How to Re-Imagine the World: A Pocket Guide for Practical Visionaries is, like Fesmire’s John Dewey and Moral Imagination and Johnson’s Morality for Humans, a book about moral imagination and its importance for our decisions about how to live. Unlike either of those books, though, there is not a whit of philosophical theorizing about ethics in this book. As Weston puts it in his opening words, “This book is a guide to creative thinking in service of radical social transformation” (p. 1). And that it is. In its small 142 pages it covers a wide variety of exhortations, strategies, and tips for creative thinking about our lives and our world.

Weston is relentlessly positive, perhaps even Romantic, without coming across as utopian in an out-of-touch, unconstrained idealism sort of way. The emphasis of the book is on vision and ideas. This is not a guide for practical activism on the ground, though there are tips and suggestions here and there about how to bring visions and ideas to life. No doubt, those truly hardened by Realpolitik and incrementalism, and many others besides, would find this book hokey and unrealistic. I see it as an invitation to be perceptive about the actual, and bring the imagination to bear in order to discover the possible. One of the best parts of the book is that it is constantly fleshing out its advice with examples that are fleshed out enough to be suggestive, but brief and sketchy enough so that you don’t get bogged down in the details. Again, some may find the style off-putting, but I liked it.

The book is organized into 23 very punch chapters (plus an exhortatory introduction and a brief afterword). The chapters are organized into five sections that are labeled with a few sentences describing major themes or strategies rather than section headings. The first section deals with the importance of working from a big-picture vision of the social reality one is aiming for. Visions inspire, bring people together, give us something positive to strive for, rather than merely a problem to fight against. The second section deals with strategies for creative or generative thinking, different ways of coming up with creative ideas that could drive social change. The third section goes in a somewhat more practical direction, providing suggestions for how to look for “tipping points” and other dynamics that make radical change achievable. Here I was reminded of Fesmire’s discussion of moral imagination as “creatively tapping a situation’s possibilities.”

The fourth section of the book discusses the way that cultural institutions, practices, norms, and world views can act as barriers to change, but also can be shifted and reconstructed to make change possible. The final section deals with more practical issues of how to build momentum behind idea, make alliances, and achieve goals. (This, for me, was the least satisfying part of the book, in part because it didn’t seem to disclose very concrete possibilities for change.)

As an aside, I think there is a bit of the book where Weston neatly captures, in plain, easily digested language, the valuable insight from Martin Heidegger’s “The Question Concerning Technology.” (At least, the take-away that I find valuable enough to keep teaching the essay.)

Let us adopt the term Resourcism for the worldview (assumption or ideology) that the Earth is essentially just a set of resources for our use. Trees are just standing lumber; rocky hills are just gravel pits waiting to happen; other animals are just living meat, and even you and I may be reduced to customers or ‘Human Resources.’ Take all of this to its logical conclusion and of course we have environmental crisis, an Earth being sucked dry. But it is hard to recognize the root cause when we don’t have a word for it Resourcism shows this world-view for what it is: a relentless, single-minded, narrow value-system, and certainly not the only one possible. (p. 115)

In sum, there are a few good reasons to read this book. If you’re in need of a boost to your moral imagination, if you’re hungry for social change but dissatisfied with the binary, oppositional terms of contemporary debate, or if you want to teach your students to think creatively about ethics and political philosophy, in terms of possibilities rather than abstract theories and criteria. Don’t expect a lot of philosophical depth or intellectual rigor, though; that’s not what this book is trying to do.

Notes on Mark Johnson’s Morality for Humans

Posted on by
7

Mark Johnson‘s Morality for Humans: Ethical Understanding from the Perspective of Cognitive Science is another book, like Fesmire’s John Dewey and Moral Imagination, that I am hoping to use as a background for Science and Moral Imagination. In the book, Johnson provides a broadly naturalistic, Deweyan pragmatist account of morality centered on moral deliberation and the role of imagination in moral deliberation. Johnson’s book complements Fesmire’s in much the way that I was hoping, i.e., it makes good use of Dewey’s distinction between “valuing” and “valuation” (though I thought Dewey called the latter “evaluation” — something I’ll have to check), and it discusses at length the relationship between science and ethics, not only the influence of science on ethics, which is a central part of Johnson’s story, but also the sense in which moral deliberation is a kind of empirical inquiry.

Johnson is probably most well-known for his long-term collaboration with cognitive linguist George Lakoff, and their work on embodied metaphor theory in cognitive semantics and its philosophical implications. It is no surprise that Johnson so adeptly reviews the empirical literature and draws implications for our understanding of morality. Johnson does not limit himself to embodied metaphor theory, but draws on the affective neuroscience of Damasio, the moral psychology of Haidt, the neurophilosophy of the Churchlands, the feminist developmental psychology of Carol Gilligan, and many other scientific sources as well as philosophers’ insights from a variety of traditions, in a way that is satisfying and provocative without becoming reductionistic or scientistic.

Chapter 1, “Moral Problem-Solving as an Empirical Inquiry,” provides a powerful argument against the idea that there is a special realm of “moral experience” and against the Kantian idea that there is a peculiar kind of “moral judgment” distinct from our ordinary repertoire of problem-solving strategies. Chapter 2 canvasses the various sources of our values, including biology, kinship, social institutions, and cultural sources. Johnson points out that some values will be universal or near-universal simply due to the necessities biological functioning and the requirements of any functioning social interaction or institution, though there will also be a lot of cultural variation. In this chapter, Johnson builds on and critically assesses Damasio’s affective neuroscience and Haidt’s moral foundations theory.

Chapter 3 reviews the popular two-process account of moral psychology, which posits an intuitive-level, affect-driven process of moral evaluation that rules most of our moral lives, and a process of moral reasoning whose main function is post-hoc justification of intuition. Johnson argues for either a third process, or another version of the second process of moral reasoning, which he calls moral deliberation, and the theory of moral deliberation occupies Chapters 4-5. Johnson does not dispute dual-process theory per se, but argues that there is another important process in our moral lives that it ignores.

Johnson’s account of moral deliberation is fully Deweyan. Moral deliberation is problem-solving inquiry that addresses a particular situation in which our habits, desires, and values are inadequate to the conditions of the particular situation. It involves gathering information about the situation and dramatic rehearsal in imagination of various possible courses of action. Johnson adopts whole hog Dewey’s view that this process is regulated by qualitative considerations, and that the goal of inquiry is to transform a situation characterized by an indeterminate, perplexing, problematic quality to one that is determinate, stable, allowing us to move forward in a satisfactory way. The process of moral deliberation as inquiry is “reasonable” if it actually transforms the situation in a way that resolves the problem or perplexity that occasioned deliberation. This process changes not only our values and our perception of the world, but the world itself and ourselves via a new structure of activities and interactions. (If anything from Dewey’s view of inquiry is missing here, it is his emphasis on “experimental testing” prior to judgment, but I must admit I am also unclear how this would work in the case of moral deliberation.)

The last three chapters held relatively less interest for me, partly because I didn’t need to be convinced of most of these things, partly because it doesn’t serve my needs as much. Chapter 6 takes to task those moral psychologists who have been tempted to revive talk of a separate “moral faculty.” Chapter 7, “Moral Fundamentalism is Immoral,” takes on both religious and rationalistic forms of moral fundamentalism, taken as the idea that there are either universally binding moral laws or absolute and foundational moral facts. These views are both incompatible with our cognitive machinery and detrimental to the needs of genuine moral deliberation, and so both impossible for humans to use and immoral insofar as we try. Moral realism is treated as an absolutist, foundationalist belief in moral facts independent of the natural picture of the world. (While this is certainly a common view under the heading of “Moral Realism,” I think Johnson is mistaken to treat this as the only way one could be a moral realist.)

Chapter 8 discusses the nature of moral experience and the moral self. Continuing the metaphysical discussion from the previous chapter, Johnson posits a pragmatist process metaphysics about values and norms, in contrast to both the objectivist metaphysics according to which values or principles are discovered, and a relativist metaphysics according to which they are arbitrarily made up. According to the pragmatist process view the metaphor of “creative transformation of our experience” and the moral deliberator as artist are much more apt. Johnson also defends Dewey’s view, which I have always found puzzling, that the ultimate end of moral deliberation is growth, on the grounds that moral deliberation requires a willingness to revise the values and habits that constitute the self to deal with ever new situations. This chapter also includes a detailed, and fairly satisfying, example of moral deliberation, about the ethics of gay marriage.

I think there is one major missed connection in Johnson’s account that connects very closely with my own interests. On the one hand, Johnson appears to hold a basically realist (if critical and fallibilist) attitude towards the science he relies on in his account. On the other hand, he denies moral realism because it is supposedly absolutist and foundationalist in untenable ways. However, Johnson himself denies that there are distinctive types of experience and inquiry. It is the first major argument of the book. Presumably, this would require us to reject the dichotomy between scientific and moral experience and inquiry, and to see his pragmatist process metaphysics as applying broadly to human knowledge, not just to values and norms. If this still permits a realist attitude about science, which I think it does, why can it not permit a realist attitude about the valuations that result from reasonable processes of moral deliberation?

My notes on Mark Johnson’s Morality for Humans

Update: I had a brief email exchange with Mark Johnson about the question at the end of the post, and he basically agrees with me. He just avoids talk of “realism” in this context because it has become associated with problematic views that are absolutist, fundamentalist, and foundationalist.

The Foundations and Scope of the Argument from Inductive Risk: An Exchange with Joyce Havstad

Posted on by
4

Introductory note: One of the most exciting parts of my work over the last couple of years has been my collaboration with Joyce C. Havstad on the science and politics of climate science. We have a paper forthcoming in Perspectives on Science, a chapter in an edited collection responding to the “Pragmatic-Enlightened Model” of science advising that influenced WG3 of the IPCC, and another journal article under review. I made some edits to one of our papers based on the content of Heather Douglas‘s Descartes Lectures and some conversations I had with Heather around the lectures, and it prompted the following exchange of ideas about and interpretations of the Argument from Inductive Risk (AIR).

Joyce C. Havstad: I’d be interested to hear more about your updated understanding of the argument from inductive risk—especially, what the difference between the argument “not applying” and “not being salient” is. I don’t want to dispute those changes to how the scope of the argument is presented in this version of the paper, but I would like to get a better sense of what that difference signifies.

Matthew J. Brown: So, here’s how I used to understand the argument from inductive risk (simplified to the case of hypothesis acceptance):

  1. Scientists make choices about whether to accept or reject hypotheses.
  2. Evidence, logic, and epistemic values leave greater or lesser amounts of uncertainty about a hypothesis.
  3. When that uncertainty is non-negligible, we have high to set our standards of acceptance.
  4. How high we set our standards of acceptance trades off false-positive and false-negative errors, all else being equal.
  5. Sometimes there are socially/ethically significant consequences of those errors.
  6. Sometimes, those consequences can be anticipated.
  7. When (2-5) hold, we must make value judgments about standards of acceptance.

That could probably be a bit more precise, but that’s basically my understanding. Under these conditions, the AIR applies if uncertainty is non-negligible, if there are social consequences, and if they can be anticipated, and it doesn’t apply when any of those are false.

Here’s my new understanding (again, simplified), which I think is a much clearer, stronger view:

  1. Scientists make choices about whether to accept or reject hypotheses.
  2. Evidence, logic, and epistemic values tell us the strength of evidential support for a hypothesis, but there is always an inductive gap between the evidence and the hypothesis.
  3. The decision to accept, infer, assert, or endorse a (non-trivial, ampliative/inductive) hypothesis is an action that requires us to “step across” that gap.
  4. No amount or strength of support necessarily compels us to assert, infer, etc.
  5. Instead, we require some sort of practical reason (i.e., values) concerning sufficiency conditions for asserting, inferring, etc.
  6. Where there are foreseeable consequences of error, these are among the relevant practical reasons.

On this interpretation, the AIR always applies. Determination of what counts as “negligible” error is already a value-laden affair. But when the evidential support for/against a hypothesis is very strong, and there don’t seem to be foreseeable socially-relevant consequences, then the AIR is not very salient. Or perhaps it would be better to say that cognitive values, bon sens, and whimsey rather than social and ethical values are salient.

This is how I interpret Douglas’s latest & greatest presentation of the AIR. What do you think?

JCH: About the old and the new inductive risk arguments: those two arguments seem quite different to me. Most importantly, it seems to me as though they would each require very different things in the way of support.

Although I think that I can see how your prior interpretation of AIR is supported by work already done—especially, for instance, by the case detailed in Douglas’s 2000 paper on inductive risk—I’m not sure I’m aware of work that supports the updated AIR.

Premise (3) in the second argument seems particularly new and interesting, and seems to require further support. I’d also want to know about the intended scope of premises (4) and (5), and to see the support for those scoped claims.

MJB: Here are some chunks from Heather’s Descartes Lectures that I take to support the new interpretation. (Whether this is sufficient to establish the point or coheres with the prior work, I’m not entirely sure, though it coheres nicely with my own intuitions about assertion.)

To upend the value-free ideal, and its presumptions about the aim of purity and autonomy in science, one needs to tackle the ideal qua ideal at the moment of justification. This is the strength of the argument from inductive risk. It points to the inferential gap that can never be filled in an inductive argument, whenever the scientific claim does not follow deductively from the evidence (which in inductive sciences it almost never does). A scientist always needs to decide, precisely at the point of inference crucial to the value-free ideal, whether the available evidence is enough for the claim at issue. This is a gap that can never be filled, but only stepped across. The scientist must decide whether stepping across the gap is acceptable. The scientist can narrow the gap further with probability statements or error bars to hedge the claim, but the gap is never eliminated.

…

Note that while [epistemic values] are very helpful in assessing the strength of the available evidence, they are mute on whether the available evidence is enough, on whether the evidence is strong enough to warrant acceptance by scientists. Epistemic values do not speak to this question at all. They help to organize and assess how strong the evidence is, but not whether it is strong enough (as, recall, it will never be complete).

…

Social and ethical values, however, do help with this decision. They help by considering the consequences of getting it wrong, of assessing what happens if it was a mistake to step across the inductive gap—i.e, to accept a claim—or what happens if we fail to step across the inductive gap and we should. In doing so, such values help us assess whether the gap is small enough to take the chance. If making a mistake means only minor harms, we are ready to step across it with some good evidence. If making a mistake means major harms, particularly to vulnerable populations or crucial resources, we should demand more evidence. Social and ethical values weigh these risks and harms, and provide reasons for why the evidence may be sufficient in some cases and not in others.

JCH: Here’s the crux of the issue as I currently see it:

Say I’m looking at a petri dish with, as I count them, 5 nematodes in it. It is true that there will always be an inductive gap that exists here: a gap between (a) my looking at the dish and thinking I have some strong evidence using my eyes and my counting ability for thinking there are 5 nematodes in it, and (b) my making the decision that the evidence provided by my eyes and my counting ability is sufficient for me to mark down that the dish has 5 nematodes in it.

And we could say that the AIR risk always applies, even to moments like the one described above, because of the presence of the inductive gap. If we go that route, then the nematode-counting case is probably just one of those cases where making a mistake risks only very minor harms, and so we’re ready to step across the gap with just the evidence of my eyes and my counting ability. On this view, we could say that the nematode-number-marking decision is a value-laden one that requires considering not just epistemic or cognitive but also ethical and social values. But surely this decision will not require nearly the same degree of involvement of non-epistemic values, consideration of risks, engagement with stakeholders, etc. that, say, the EPA’s decision about where to set the acceptable levels of dioxin regulation did, or the IPCC’s decision to offer a set of three particular global temperature increase pathways should. Despite the AIR applying in all three cases (on this interpretation), the cases will not be ethically and socially value-laden in the same ways or to nearly the same extent.

Alternatively, we could maintain something of a distinction between the notion of an omnipresent inductive gap and the idea of inductive risk. If we go this route, then it is true that the nematode-counting case includes, as always, an inductive gap; but it is not necessarily true that the nematode-number-marking decision is an inductively risky one (again, because it is probably just one of those cases where making a mistake risks only very minor harms, in the sense that any decision ever risks very minor harms). On this view, the AIR applies only to a particular set of the decisions involving the inductive gap—for instance, those in which there are notable, foreseeable consequences of error with significant ethical and social implications. And probably also those cases which might have such implications but where the consequences are not as foreseeable (i.e., the so-called “gray areas”). Here (on this interpretation), whether and how the AIR applies tracks whether and how the relevant cases will be significantly ethically and socially value-laden.

Either way, not all cases with an inductive gap are the same with respect to their ethical and social value-ladenness. I think that I care less about being able to say that all decisions are ethically and socially value-laden (in what looks to me like a pretty trivial sense), than I do about being able to identify which decisions are significantly ethically and socially value-laden (in a discriminating and useful sense). This is because I want to be able to identify and address those extremely risky decisions which are currently being made without proper consideration of ethical and social values, but which are in dire need of them—like the EPA and the IPCC cases, but not like the nematode-counting one. To me, it is a strength of your prior interpretation of the AIR that it is able to clearly discriminate amongst cases in this way; the newer interpretation looks to be somewhat weakened along this dimension, though that may be the result of some generalization or vagueness in this [i.e., MJB’s] rough draft of the argument.

Regardless: whether we want to say that the AIR always applies, or that it is merely the inductive gap which is always present, I think that it is clear that not all decisions to cross the inductive gap are the same in terms of value-ladenness. Some are much, much riskier than others; and some require the consideration of ethical and social values to a far greater extent and perhaps even in a different kind of way than others.

What all this means is that I don’t think we can reliably infer, merely from the presence of an inductive gap, that we are in one of these situations rather than another. In other words, it’s not the inductive gap itself which carries the relevant ethical and social entailments which concern me; I care about the relevant social and ethical entailments; so the mere presence of an inductive gap does not for me a relevant case make. And (so my thinking goes), we ought not to treat it like it does.

MJB: Yes, I agree that not all decisions to cross the inductive gap are the same, in terms of value-ladenness. But is the difference between the cases primarily an epistemic question or primarily a values question? In other words, are some decisions less value-laden as such, or are the values just less significant in some cases?

I think on my old interpretation, it is natural to see the question as primarily an epistemic one. Inductive risks are a worry when risks of error are high, which requires uncertainty. Lower uncertainty, lower risk of error, less worry about IR. I think this opens up the AIR to the problems with “the lexical priority of evidence” that I raise in “Values in Science beyond Underdetermination and Inductive Risk.”

On the new interpretation, the difference is primarily an ethical one. Inductive risks are a worry when risks of error are salient, which requires social consequences to be foreseeable and significant. Stronger evidence reduces our worry about error, but only if it is strong enough. In some areas, social/ethical implications may be weak or may not exist, but we still need some kind of values to license making the inference/assertion. Maybe they’re merely pragmatic/aesthetic rather than social/ethical. (Here I’m thinking about Kent Staley‘s work on the AIR and the Higgs discovery, which shows that IR is an issue even when social and ethical values really aren’t, except maybe the about of money spent on the LHC.)

Also, I think that on this view, I think we can see why the direct/indirect roles distinction has merit but needs to be reconfigured and treated as defeasible. (But that’s a promissory note on an argument I’m trying to work out.)

I also think there is strategic value in insisting that the AIR applies everywhere, and that all the decisions in science are potentially value-laden. Scientists are too quick to dismiss potential ethical concerns and to see their work as governed mainly by technical/epistemic issues, and they are not encouraged to work very hard to foresee possible consequences of their decisions. They often don’t even realize they’re making decisions. And while the social/ethical consequences in some cases are quite obvious, there are plenty of cases where they crop up where least expected. So I’d rather have working hard to foresee the possible consequences of seemingly technical decisions be a core part of the job description, rather than thinking of it as an exceptional case. (This is partly why I’m currently focusing on moral imagination as a central concept for the values in science debate.)

JCH: I think I agree with most everything you say here, especially the part about the AIR being about not just error and uncertainty but also about risk and consequences. However, I also see both those things as being well represented in your prior interpretation; I might even find them less well represented in the new one.

Perhaps the new interpretation does more to highlight the ubiquity of the phenomenon under study. However, when the argument is glossed in that way (as it is, for instance, in your final paragraph), I have a hard time distinguishing the supposed problem of inductive risk from the plain old problem of induction.

BTW, I’ve been pondering the scope of the AIR for quite some time now, so I’m very pleased to be going back and forth on this issue with you now. At the very least I’m starting to better understand the nature of and motivation for the ubiquity claim, even if I’m not quite persuaded of it.