Brains
On Philosophy of Mind and Related Matters
On Philosophy of Mind and Related Matters
Brains
The Journal of Visualized Experiments
Do Philosophers Rank their Journals?
A while ago Mark Couch alerted me to an article by Jeffrey Di Leo in Inside Higher Ed. Di Leo's thesis is this:
"It is one sign of the good health of the humanities that they have not caught rank and brand fever like many of the other disciplines in the American academy. Whereas one can readily find rankings of science or business journals, there is silence when it comes to rankings of humanities journals."
Although Di Leo makes some good points—e.g., that it is difficult to rank together journals that specialize in different areas of philosophy and that good work may be published in lesser known journals—his main premise is completely backwards. Most surprisingly, his main example of a humanity discipline is philosophy!
I can't speak about other humanities. But philosophers definitely do rank their journals (here is a recent example) and more importantly, they pay a lot of attention—sometimes too much attention—to where their work is published.
Contrary to what Di Leo argues or implies, my experiences suggests that by and large, within mainstream philosophy departments decisions about awards, grants, hires, and promotions are heavily influenced by where people publish. The "better" the journals where someone has published, the better her chances to be hired, promoted, and given grants or awards.
Ranking journals and paying so much attention to where someone publishes creates some distortions, of course. Where someone's work is published is at best a coarse and somewhat unreliable measure of its quality, yet people often take that shortcut (which in many cases is a necessary shortcut for lack of time or expertise). Furthermore, acceptances at the best journals are almost certainly biased at least somewhat in favor of people who work at the best philosophy departments (in spite of double blind refereeing, where it is done).
Nevertheless, journal rankings are useful if used with the proverbial grain of salt. Publishing in, say, Phil Review doesn't make anyone a genius, and publishing in the Journal of Inferior Philosophy doesn't make anyone an inferior philosopher. But a journal will not be ranked high forever if its quality declines. Witness that many philsophers now consider J. Phil (which ten years ago was still considered #1 or at worst #2) as lower than Phil Review, Nous, and Philosophers' Imprint. That being the case, whether a paper is published in a good journal is a very useful indicator of its likely quality—not to mention, of how many people will read and cite it.
"It is one sign of the good health of the humanities that they have not caught rank and brand fever like many of the other disciplines in the American academy. Whereas one can readily find rankings of science or business journals, there is silence when it comes to rankings of humanities journals."
Although Di Leo makes some good points—e.g., that it is difficult to rank together journals that specialize in different areas of philosophy and that good work may be published in lesser known journals—his main premise is completely backwards. Most surprisingly, his main example of a humanity discipline is philosophy!
I can't speak about other humanities. But philosophers definitely do rank their journals (here is a recent example) and more importantly, they pay a lot of attention—sometimes too much attention—to where their work is published.
Contrary to what Di Leo argues or implies, my experiences suggests that by and large, within mainstream philosophy departments decisions about awards, grants, hires, and promotions are heavily influenced by where people publish. The "better" the journals where someone has published, the better her chances to be hired, promoted, and given grants or awards.
Ranking journals and paying so much attention to where someone publishes creates some distortions, of course. Where someone's work is published is at best a coarse and somewhat unreliable measure of its quality, yet people often take that shortcut (which in many cases is a necessary shortcut for lack of time or expertise). Furthermore, acceptances at the best journals are almost certainly biased at least somewhat in favor of people who work at the best philosophy departments (in spite of double blind refereeing, where it is done).
Nevertheless, journal rankings are useful if used with the proverbial grain of salt. Publishing in, say, Phil Review doesn't make anyone a genius, and publishing in the Journal of Inferior Philosophy doesn't make anyone an inferior philosopher. But a journal will not be ranked high forever if its quality declines. Witness that many philsophers now consider J. Phil (which ten years ago was still considered #1 or at worst #2) as lower than Phil Review, Nous, and Philosophers' Imprint. That being the case, whether a paper is published in a good journal is a very useful indicator of its likely quality—not to mention, of how many people will read and cite it.
Can You Make a Conscious Machine?
I am writing a thesis on the possibility of machine consciousness (e.g., the possibility of creating a silicon-based system that has subjective, qualitative experiences in the same way we take ourselves to have them). In my informal (and very limited) polling, it seems that many philosophers are sympathetic to the project.
However, not all functions can be realized by all structures (e.g., you can’t make a car out of paper towels). Given the many fundamental asymmetries in structure between humans and silicon-based systems, it is at least plausible that human consciousness cannot be realized in silicon-based structures.
Of course, proponents of machine consciousness will reply that the properties relevant to consciousness are multiply realizable, but do they have any evidence?
I was wondering what others take to be the strongest arguments either for or against the possibility of machine consciousness. What is the strongest reason to think human consciousness is realizable in silicon? And what is the strongest reason to think it is not? Centrally, what aspect(s) of a putative realizer seem most relevant to realizing human-like consciousness?
However, not all functions can be realized by all structures (e.g., you can’t make a car out of paper towels). Given the many fundamental asymmetries in structure between humans and silicon-based systems, it is at least plausible that human consciousness cannot be realized in silicon-based structures.
Of course, proponents of machine consciousness will reply that the properties relevant to consciousness are multiply realizable, but do they have any evidence?
I was wondering what others take to be the strongest arguments either for or against the possibility of machine consciousness. What is the strongest reason to think human consciousness is realizable in silicon? And what is the strongest reason to think it is not? Centrally, what aspect(s) of a putative realizer seem most relevant to realizing human-like consciousness?
Knowing That P Without Believing That P
Abstract:
The standard view in contemporary epistemology is that knowledge entails belief. Proponents of this claim rarely offer a positive argument in support of it. Rather, they tend to treat the view as obvious, and if anything, support the view by arguing that there are no convincing counterexamples. We find this strategy to be problematic. In particular, we do not think the standard view is obvious, and moreover, we think there are cases in which a subject can know some proposition P without (or at least without determinately) believing that P. In accordance with this, we present four plausible examples of knowledge without belief, and we provide empirical evidence which suggests that our intuitions about these scenarios are by no means atypical.
Comments are welcome.
[Cross-posted at Experimental Philosophy and The Splintered Mind]
Aizawa on Turing-Equivalent Computation and Cognitive Science
In a previous thread, Ken Aizawa suggests that I'm insufficiently pluralistic about computation in cognitive science and to substantiate his criticism he points to his forthcoming article "Computation in Cognitive Systems; It's not al about Turing-Equivalent Computation" (available on his website).
Having read Ken's nice paper, I only have time for a few quick comments.
1. Ken correctly points out that there are several notions of computation. (I make the same point in a paper that he refers to.)
2. Ken correctly points out that many people including myself think there is something special and theoretically deep about what he calls Turing-equivalent computation, by which he seems to mean the kind of computation that can be performed by Turing machines (computation of Turing-computable functions). They're right, because in fact this is the core theoretical notion of computation, with lots of deep results about it.
3. Ken correctly points out that the notion of Turing-computable functions can be generalized to study functions of natural numbers (or equivalently of strings of letters from a finite alphabet) that are not computable by Turing machines. This enterprise was started by Turing himself and is a large branch of computability theory. (Anyone who takes a nontrivial course in computability theory knows this.) But contrary to what Ken seems to suggest, the study of functions that are uncomputable by Turing machines is not based on a different notion of computation from that of Turing machines—it's the very same notion; in fact, the whole subject matter is defined in terms of functions that are like those computable by Turing machines but cannot be computed by Turing machines.
4. Ken persuasively argues that Turing machines and the related notion of computabiltiy probably played only a minor role in McCulloch's thinking at the time he wrote his 1943 paper with Pitts. But Ken seems to underestimate the theoretical significance of computation-theoretic results in characterizing the power of McCulloch-Pitts nets and other neural networks. (The latter obviously is not discussed in the 1943 paper.) Ken also seems to underestimate the important role that the connection between McCulloch-Pitts nets and Turing-computability played in the history of cybernetics and cognitive science. For the beginning of an account of that history, based on extensive archival research, see Chapters 5 and 6 of my Ph.D. dissertation.
5. Ken asserts that the notion of "computed vs. uncomputed cortical maps" deployed by some neuroscientists "is not a Turing—equivalent form of computation" (p. 17). But I didn't notice anything in Ken's paper that determines what relationship there is or isn't between the notion of computation deployed in this area of neuroscience and Turing-computabilitiy, so I don't know why Ken is so confident in his assertion.
6. Ken also points out that often cognitive scientists talk about computation as symbol manipulation or digital symbol manipulation, without mentioning the kind of "finitude constraints" that are important to Turing-computability. This is true but doesn't mean that the finiteness constraints are not implicitly assumed to be in place (except by people like Jack Copeland); after all the brain has only a finite number of neurons etc.
7. Ken's pluralism seems to be based on something like the following argument: if scientist A uses "computation" in pursuit of goal X and scientist B uses "computation" in pursuit of goal Y and X is different than Y, than scientists A and B use two different notions of computation. This is a fallacy. Maybe there are two different notions of computation, maybe they aren't. It takes a lot more than this to show that two notions of computation are the same or different. More generally, it takes a lot of theoretical work to compare and contrast different notions of computation and see how they relate to one another. That's why, contrary to what Ken suggests, it's very helpful to have an umbrella notion of computation, of which other notions (including all those mentioned by Ken) are species.
8. In conclusion, reading Ken's paper convinced me that I have just just the right kind of pluralism for sorting out competing claims about computation in cognitive science.
Having read Ken's nice paper, I only have time for a few quick comments.
1. Ken correctly points out that there are several notions of computation. (I make the same point in a paper that he refers to.)
2. Ken correctly points out that many people including myself think there is something special and theoretically deep about what he calls Turing-equivalent computation, by which he seems to mean the kind of computation that can be performed by Turing machines (computation of Turing-computable functions). They're right, because in fact this is the core theoretical notion of computation, with lots of deep results about it.
3. Ken correctly points out that the notion of Turing-computable functions can be generalized to study functions of natural numbers (or equivalently of strings of letters from a finite alphabet) that are not computable by Turing machines. This enterprise was started by Turing himself and is a large branch of computability theory. (Anyone who takes a nontrivial course in computability theory knows this.) But contrary to what Ken seems to suggest, the study of functions that are uncomputable by Turing machines is not based on a different notion of computation from that of Turing machines—it's the very same notion; in fact, the whole subject matter is defined in terms of functions that are like those computable by Turing machines but cannot be computed by Turing machines.
4. Ken persuasively argues that Turing machines and the related notion of computabiltiy probably played only a minor role in McCulloch's thinking at the time he wrote his 1943 paper with Pitts. But Ken seems to underestimate the theoretical significance of computation-theoretic results in characterizing the power of McCulloch-Pitts nets and other neural networks. (The latter obviously is not discussed in the 1943 paper.) Ken also seems to underestimate the important role that the connection between McCulloch-Pitts nets and Turing-computability played in the history of cybernetics and cognitive science. For the beginning of an account of that history, based on extensive archival research, see Chapters 5 and 6 of my Ph.D. dissertation.
5. Ken asserts that the notion of "computed vs. uncomputed cortical maps" deployed by some neuroscientists "is not a Turing—equivalent form of computation" (p. 17). But I didn't notice anything in Ken's paper that determines what relationship there is or isn't between the notion of computation deployed in this area of neuroscience and Turing-computabilitiy, so I don't know why Ken is so confident in his assertion.
6. Ken also points out that often cognitive scientists talk about computation as symbol manipulation or digital symbol manipulation, without mentioning the kind of "finitude constraints" that are important to Turing-computability. This is true but doesn't mean that the finiteness constraints are not implicitly assumed to be in place (except by people like Jack Copeland); after all the brain has only a finite number of neurons etc.
7. Ken's pluralism seems to be based on something like the following argument: if scientist A uses "computation" in pursuit of goal X and scientist B uses "computation" in pursuit of goal Y and X is different than Y, than scientists A and B use two different notions of computation. This is a fallacy. Maybe there are two different notions of computation, maybe they aren't. It takes a lot more than this to show that two notions of computation are the same or different. More generally, it takes a lot of theoretical work to compare and contrast different notions of computation and see how they relate to one another. That's why, contrary to what Ken suggests, it's very helpful to have an umbrella notion of computation, of which other notions (including all those mentioned by Ken) are species.
8. In conclusion, reading Ken's paper convinced me that I have just just the right kind of pluralism for sorting out competing claims about computation in cognitive science.
Information Processing, Computation, and Medium Independence
In response to a previous thread, Jonathan Livengood asked some very good questions about, roughly, what should count as information processing and computation in physical systems. Perhaps it will help to take a step back.
In my early work on computation, I argued that, roughly, only physical processes that take strings of digits as inputs and return strings of digits as outputs by following a rule defined over the inputs (and possibly internal states) count as computing systems . My reason had to do with the centrality of the formalisms of computability theory to the notion of computation. As to analog computers, which do not manipulate strings of digits but are still called computers, I argued that they are "computational" only by courtesy and for contingent historical reasons .
I later realized that, although there was something right about my early purism about computation, it was unhelpful to try to restrict the notion of computation to digital computation in the face of important yet broader uses of the term "computation" in many sciences, including neuroscience. (BTW, I had this realization in time to dodge Ken Aizawa's criticism that I was insufficiently pluralistic; Ken's criticism does apply to my former self, though.).
After that, I needed to characterize a notion of computation more general than that of digital computation (without appealing to representation, of course, otherwise I would have gone against one of my core views about computation ).
What came to my rescue is the notion of medium independence. Medium independence was introduced by Justin Garson in his 2003 MA thesis, part of which was published in a beautiful and underappreciated article in Philosophy of Science on "The Introduction of Information in Neurobiology ".
Justin pointed out that the first person to talk about neural systems transmitting information was Edgar Adrian (1928), on the grounds of his groundbreaking discovery of some crucial properties of neural signals ("all or none", "rate coding", and "adaptation"). Justin reconstructed Adrian's notion of information as involving medium independence:
"Medium independence: The structure S—for example, the structure relation that obtains between the units of a sequence of action potentials—can be instantiated across a wide range of physical mechanisms." (Garson 2003, p. 927)
While Justin's medium independence is not necessary for carrying (natural) information in the usual sense, a slightly modified version of it seems well suited for characterizing the general notion of computation. So in my more recent work, I characterize computation in the generic sense as (roughly) the functional manipulation of medium independent vehicles according to rules, where a variable is medium independent just in case it is manipulated on the grounds of similarities and differences between its parts along a certain dimension of variation, irrespective of its more concrete physical properties.
Example 1: various sensory receptors transduce all kinds of physical variables into spike trains, which are then conveyed to the nervous system, which in turn manipulates these spike trains. This was one of Adrian's amazing discoveries: neural fibers carry the same kind of signals regarless of their physical inputs. Thus, neural processes are computations in the generic sense.
Example 2: computers manipulate strings of digits, which are well defined so long as there are distinguishable types and an ordering relation, regardless of the details of their physical implementation. The same digital computation can be performed in mechanical, electronic, electromechanical, etc. media. Thus, the activity performed by digital computers are computations in the generic sense (and digital computation is a species of computation in the generic sense).
Example 3: mutatis mutandis, analog computers manipulate their own type of medium independent vehicles. Thus, analog computation is a kind of generic computation.
I hope someone can see the beauty of this. We now have an explicit, non-semantic characterization of computation in general, of which analog computation, digital computation, etc. are species. Thanks Justin!
Anyway, then there is the question of information processing. Obviously information can be "processed" in a medium dependent way, as done by the Watt governors or float regulators that we've been discussing. If Jonathan insists in calling this type of thing information processing, so be it. But information processing can also be done in a medium independent way, and IMO that's what most people mean when they talk about information processing. In any case, if you care about what does and does not compute it's important to notice the difference between the two cases, because medium-dependent information processing does not entail computation whereas medium-independent information processing does entail computation.
In my early work on computation, I argued that, roughly, only physical processes that take strings of digits as inputs and return strings of digits as outputs by following a rule defined over the inputs (and possibly internal states) count as computing systems . My reason had to do with the centrality of the formalisms of computability theory to the notion of computation. As to analog computers, which do not manipulate strings of digits but are still called computers, I argued that they are "computational" only by courtesy and for contingent historical reasons .
I later realized that, although there was something right about my early purism about computation, it was unhelpful to try to restrict the notion of computation to digital computation in the face of important yet broader uses of the term "computation" in many sciences, including neuroscience. (BTW, I had this realization in time to dodge Ken Aizawa's criticism that I was insufficiently pluralistic; Ken's criticism does apply to my former self, though.).
After that, I needed to characterize a notion of computation more general than that of digital computation (without appealing to representation, of course, otherwise I would have gone against one of my core views about computation ).
What came to my rescue is the notion of medium independence. Medium independence was introduced by Justin Garson in his 2003 MA thesis, part of which was published in a beautiful and underappreciated article in Philosophy of Science on "The Introduction of Information in Neurobiology ".
Justin pointed out that the first person to talk about neural systems transmitting information was Edgar Adrian (1928), on the grounds of his groundbreaking discovery of some crucial properties of neural signals ("all or none", "rate coding", and "adaptation"). Justin reconstructed Adrian's notion of information as involving medium independence:
"Medium independence: The structure S—for example, the structure relation that obtains between the units of a sequence of action potentials—can be instantiated across a wide range of physical mechanisms." (Garson 2003, p. 927)
While Justin's medium independence is not necessary for carrying (natural) information in the usual sense, a slightly modified version of it seems well suited for characterizing the general notion of computation. So in my more recent work, I characterize computation in the generic sense as (roughly) the functional manipulation of medium independent vehicles according to rules, where a variable is medium independent just in case it is manipulated on the grounds of similarities and differences between its parts along a certain dimension of variation, irrespective of its more concrete physical properties.
Example 1: various sensory receptors transduce all kinds of physical variables into spike trains, which are then conveyed to the nervous system, which in turn manipulates these spike trains. This was one of Adrian's amazing discoveries: neural fibers carry the same kind of signals regarless of their physical inputs. Thus, neural processes are computations in the generic sense.
Example 2: computers manipulate strings of digits, which are well defined so long as there are distinguishable types and an ordering relation, regardless of the details of their physical implementation. The same digital computation can be performed in mechanical, electronic, electromechanical, etc. media. Thus, the activity performed by digital computers are computations in the generic sense (and digital computation is a species of computation in the generic sense).
Example 3: mutatis mutandis, analog computers manipulate their own type of medium independent vehicles. Thus, analog computation is a kind of generic computation.
I hope someone can see the beauty of this. We now have an explicit, non-semantic characterization of computation in general, of which analog computation, digital computation, etc. are species. Thanks Justin!
Anyway, then there is the question of information processing. Obviously information can be "processed" in a medium dependent way, as done by the Watt governors or float regulators that we've been discussing. If Jonathan insists in calling this type of thing information processing, so be it. But information processing can also be done in a medium independent way, and IMO that's what most people mean when they talk about information processing. In any case, if you care about what does and does not compute it's important to notice the difference between the two cases, because medium-dependent information processing does not entail computation whereas medium-independent information processing does entail computation.
A Response to Machery's Response
The latest issue of BBS includes a précis of Edouard Machery's Doing Without Concepts
—the book that boldly argues that the term "concept" should be eliminated from psychology. The fourth tenet of Machery's Heterogeneity Hypothesis (HH) proposes that prototypes, exemplars, and theories
—three types of concept
—are used in distinct cognitive processes. Gualtiero Piccinini and I wrote a short response arguing that Machery has not provided enough evidence that prototypes and exemplars are used in distinct cognitive processes. If we are right, then Machery's argument for concept eliminativism as he presents it doesn't go through.
Interestingly, Machery chooses to respond to us by weakening his argument. Machery maintains that even if exemplars and prototypes are used in the same cognitive processes, theories are used in different cognitive processes from exemplars and prototypes. Thus, Machery concludes, "there are no generalizations about how concepts are used in cognitive processes" (2010, 237).
But Machery's response is less satisfying than it may appear. Our commentary focused on prototypes and exemplars because that's where Machery made his strongest case for differences in the cognitive processes. Machery offers little, if any, support for the claim that theories are used in distinct cognitive processes from prototypes and exemplars. In fact, everything Machery says is consistent with the hypothesis that theories are prototypes plus some causal information of a category (incidentally, many psychologists also believe that theories are enriched versions of prototypes). If theories are just augmented prototypes (i.e., a kind of prototype), then a fortiori they are used in the same cognitive processes. The burden is on Machery to provide evidence that theories are truly distinct from prototypes and are used in distinct cognitive processes.
Unless, of course, theories are seen as what Piccinini (forthcoming) calls "linguistic concepts".
Interestingly, Machery chooses to respond to us by weakening his argument. Machery maintains that even if exemplars and prototypes are used in the same cognitive processes, theories are used in different cognitive processes from exemplars and prototypes. Thus, Machery concludes, "there are no generalizations about how concepts are used in cognitive processes" (2010, 237).
But Machery's response is less satisfying than it may appear. Our commentary focused on prototypes and exemplars because that's where Machery made his strongest case for differences in the cognitive processes. Machery offers little, if any, support for the claim that theories are used in distinct cognitive processes from prototypes and exemplars. In fact, everything Machery says is consistent with the hypothesis that theories are prototypes plus some causal information of a category (incidentally, many psychologists also believe that theories are enriched versions of prototypes). If theories are just augmented prototypes (i.e., a kind of prototype), then a fortiori they are used in the same cognitive processes. The burden is on Machery to provide evidence that theories are truly distinct from prototypes and are used in distinct cognitive processes.
Unless, of course, theories are seen as what Piccinini (forthcoming) calls "linguistic concepts".
14TH CONGRESS OF LOGIC, METHODOLOGY AND PHILOSOPHY OF SCIENCE NANCY, JULY 19-26, 2011
I know that most folks get their conference CFPs, etc. from other sources, but this one seems to me a) to be flying a little low on the radar and b) to be pretty cool. I went to one of these a long time ago in Florence, which was both a very good conference, and a truly amazing world class city. It would be great to have a good representation of brains readers. (I've already submitted my abstracts, but don't let that deter you.)
Check it out here .
Check it out here .
A New Account of the Systematicity of Thought
In other Aizawa-relevant news, Steven Philips and Williams Wilson have a new theory of the systematicity of thought based on category theory.
With their publication, they have joined an elite group of academics who have referred to my book, The Systematicity Arguments. (Fodor mentions it in LOT 2 and McLaughlin mentions it in his "Systematicity Redux" .) Indeed, Philips and Wilson do more than just mention the book. They have an entire section of largely positive discussion some of the material. The material is probably off the radar of most philosophers in the PLoS Computational Biology, but it is freely available.
With their publication, they have joined an elite group of academics who have referred to my book, The Systematicity Arguments. (Fodor mentions it in LOT 2 and McLaughlin mentions it in his "Systematicity Redux" .) Indeed, Philips and Wilson do more than just mention the book. They have an entire section of largely positive discussion some of the material. The material is probably off the radar of most philosophers in the PLoS Computational Biology, but it is freely available.
http://www.ploscompbiol.org/I'm sure my royalties for that book will now be waaay more than ten dollars for this year.article/info:doi/10.1371/ journal.pcbi.1000858 (HTML)
http://www.ploscompbiol.org/article/fetchObjectAttachment. action?uri=info%3Adoi%2F10. 1371%2Fjournal.pcbi.1000858& representation=PDF (PDF)
http://www.ploscompbiol.org/article/fetchObjectAttachment. action?uri=info%3Adoi%2F10. 1371%2Fjournal.pcbi.1000858& representation=XML (XML)
Subtyping and Multiple Realization
As many of you may know, I have been thinking about the following problem for a while:
It has recently come to my attention that Michael Esfeld, Christian Sachse, and Patrice Soom have been developing views (roughly) along the lines of door #3. See, for example,
"Theory Reduction by Means of Functional Sub-Types"
Reductionism in the Philosophy of Science.
"Functional Subtypes" .
Esfeld and Sachse also have a book forthcoming from Routledge developing a version of door #3.
I'm reading through their stuff now, but the most jarring thing for me is that they claim that they are exploring subtyping as a logically possible thing for scientists to do. But, if this is merely a logically possible thing for them to do, and not something they actually do, then why think this has much to do with science or reduction?
All that's rough, but take this post as a trailer for this topic.
Suppose that scientists discover a high level property G that is prima facie multiply realized by two sets of lower level properties, F1, F2, …, Fn, and F*1, F*2, …, F*m. One response would be to take this situation at face value and conclude that G is in fact so multiply realized. A second response, however, would be to eliminate the property G and instead hypothesize subtypes of G, G1 and G2, and say that G1 is uniquely realized by F1, F2, …, Fn, and that G2 is uniquely realized by F*1, F*2, …, F*m. This second response would eliminate a multiply realized property in favor of two uniquely realized properties A third possible scientific strategy would be to keep G and add subtypes G1 and G2. What do scientists actually do?With Carl Gillett, I've been arguing, in essence, that scientists take door #1. This answer is defended in a forthcoming paper with Carl, "The Autonomy of Psychology in the Age of Neuroscience" ,
It has recently come to my attention that Michael Esfeld, Christian Sachse, and Patrice Soom have been developing views (roughly) along the lines of door #3. See, for example,
"Theory Reduction by Means of Functional Sub-Types"
Reductionism in the Philosophy of Science.
"Functional Subtypes" .
Esfeld and Sachse also have a book forthcoming from Routledge developing a version of door #3.
I'm reading through their stuff now, but the most jarring thing for me is that they claim that they are exploring subtyping as a logically possible thing for scientists to do. But, if this is merely a logically possible thing for them to do, and not something they actually do, then why think this has much to do with science or reduction?
All that's rough, but take this post as a trailer for this topic.
