The past five days or so have been one of those rare times in my life when my social life expands to the exclusion of all my concrete projects. It's been real good.( Read more... )
I would say that attitudes, beliefs, and feelings are not memories per se. Well, I guess, but then I'm not too sure about any distinction between memory and knowledge. If I learn that south american monkeys have a different kind of nose from old world monkeys, and I remember that, is that memory or knowledge? Similarly, if I learn to ride a bike, is that procedural memory or procedural knowledge? I'm honestly asking, is there some clear distinction that's drawn? 'Cos it seems to me that 'memory' can be stretched to any information stored in the mind that's derived from past experience, which is almost all of it. And if knowledge counts, then belief would have to count, insofar as they differ in terms of how reality is. And I think beliefs and attitudes blur into one another ('I hate him' blurs into 'he is a bastard').
And Freud was a fraud. You take that back!
if your hypothesis (that most memories have some of each), it would be much harder to explain these cases. Well fine, I guess - I'm not (yet) an empirical neurophysiologist, so I can't say much here, though for what it's worth I don't think there's too much of a incompatibility here: if any given memory can have two aspects when it's used in two ways, then it would probably be used 'by' two different loci. I suppose though it then becomes really a semantic issue (in the bad sense) whether we say that there are 'two memories' or one memory processed in two different ways.
Sorry to be slow to get back to you on your metaphysics opus No problem. Wrists are demanding mistresses.
If I learn that south american monkeys have a different kind of nose from old world monkeys, and I remember that, is that memory or knowledge?
ok, here's where psychologists would deploy the semantic v. episodic memory distinction, sort of a weak-ass attempt to separate abstractions and propositions from the garbled mass of associations which can't be distilled into simple sentences/propositions or beliefs, but still kick around in our head.
zB. - your experience when learning this fact about monkeys, filled with all sorts of extraneous information - the tone of voice that the speaker related the information to you, the brightness of your environment, the feeling of your chafing boxer-briefs- is all supposedly, imprinted into our "episodic memory" and cannot be recalled as a collection of propositions, just one baffling chaotic experience. After the fact, we then pull atomic or simple propositions out of this mess. "my boxers were chafing", "south american monkeys have a different type of nose." and so on. interestingly, in this sort of analysis, there's no ability to generalize about experience by just calling the episodic memory. the only things we can report later on and hence (?) "know" are stored in semantic memory. I envision it with the help of the analogy of a transparency used as a visual aid. We have a jumbled image below (our episodic memory) and then above we have a transparent overlay, with little boxes containing the propositions contained in our semantic memory. There should be some mechanism that leaves little nodes or anchors in the episode that connect to the propositions.
psychologists apparently don't distinguish between memory and knowledge per se. (or at least those who have informed what i've read in my low-level psych courses, i'm sure there ARE psychologists who maintain this distinction, i'd wager those of the philosopher persuasion DO) Anyway, it seems like 'knowledge', which is by nature abstracted (?), can be seen as the content of our 'semantic' store
i think, however, conceptions of knowledge and the dividing line between bodily functions/abilities and knowledge that one can articulate is an interesting topic for discussion. for example, swamis or yogis in india consider your autonomic processes 'knowledge' which most western philosophers would balk at, like woah. anyway, my roommates are drunk and boisterous and this has been incoherent enough already. this house is not conducive to thought on ANY level.
If I learn that south american monkeys have a different kind of nose from old world monkeys, and I remember that, is that memory or knowledge?
I stand corrected by my buddy.
And, after reading both your comments, and remembering some stuff I should have remembered earlier about different systems of reasoning, I'm going to change my position somewhat:
I think it would be appropriate to blur the line on procedural memory/knowledge. And I think that it is implicit in that we can't consciously access it declare it.
And I think that it is a different kind of memory (not just a different use of the same memory) than episodic and semantic memories.
I don't see how one could confirm the existence of anything like "implicit episodic memory." That sounds like Freudian bullshit to me. And no, I'm not taking that back until somebody proves me wrong.
For semantic memory, I remember now that there is a supported hypothesis for two systems of reasoning, one implicit and one explicit, which would work using something like semantic knowledge/memory. If I remember right, it was an open question whether or not it was the same knowledge being used in two ways, or two different kinds of knowledge, although I think the prof who was teaching me about it believed strongly that it was the latter.
-----
I don't know where that puts me relative to your intuitions. Really, I ought to look all this stuff up again.
But I disagree with the claim that the "two use" and "two memories" theories are identical except for semantics. Consider the episodic/procedural case. Suppose two subjects are trained in a task, and one, subject A, can remember the training but is not any better at performing the task for it, and the other, subject B, can't remember the training but is good at the task (this has happened.)
Now, either we say:
(1) A does not have the procedural memory but has the episodic memory, and the opposite is the case for B.
or
(2) A still has the combine procedural/episodic memory of the training experience, but has simply lost the procedural use of that memory. The evidence for the original memory (which could potentially have both uses) is the observable episodic memory.
The latter theory has an extra entity but no predictive power. So if only on the grounds of parsimony, it should be rejected. Moreover, it would require (given reasonable beliefs about the relationship between memory and the brain) that there be some place in brain where the combined episdoic/procedural memory is encoded that is distinct from the place where each of the two uses is encoded. I don't think I've heard any evidence to that effect.
Moreover, it would require (given reasonable beliefs about the relationship between memory and the brain) that there be some place in brain where the combined episdoic/procedural memory is encoded that is distinct from the place where each of the two uses is encoded. I don't think I've heard any evidence to that effect.
Word. that's tricky business, a distinct encoding area for the memory as opposed to the use. i dont think i buy it either.
but: here's what im thinking: (colon inside a colon, weird) episodic memory isn't productive, in the sense that (maybe due to its unruliness) episodic memories cant directly influence behavior or self-report. they have to be translated into component parts, pared down, specified to particular tasks, etc. what i refered to as 'semantic memory' on the other hand, is production ready, its the actor's formula for action, belief, thought. i dunno, it seems as though episodic memory would have to be more primal, and that as soon as you begin to really analyze it, you've started shifting it over into semantic memory, using abstract characteristics/labels/whatever to describe it.
which birngs about the question of form. my thoughts sort of presuppose that there is NOT some uniformity in encoding between the two memories, but rather a function (or suite of functions) that maps parts of episodic memory to semantic. No matter how hard i try, i cannot bring myself to buy into mentalese wholeheartedly. maybe mentalese is used for cross-compartmental correspondence, but i think it would only be as a last resort. things wouldnt be translated into mentalese unless they absolutely HAVE to be. so, episodic memory is a MESS, because its got tons of different types of inputs and coding and what not. (at the very least info from the 5 senses, and then some emotional information, at the most basic level), but it neednt be encoded into a single form or language.
but then again we're back to a problem - this procedural knowledge that one CAN'T report on, is it in mentalese? Suppose we posit that 'knowledge', which i think is just a subset of memory, is all in mentalese, as to facilitate production and transmission of information, WHERE THE HELL does the explicit/implicit demarcation fall? if 'physical' knowledge, like the autonomic system's functions, is also stored in mentalese - why CAN"T i report on it?
I don't see how one could confirm the existence of anything like "implicit episodic memory." That sounds like Freudian bullshit to me. Well, the natural suggestion would seem to be that we confirm it when people 'regain' access to memories that they previously had no access to, such as of events in childhood. That's not to deny that there aren't cases where such memories could be mere fantasies, but insofar as there is the possibility of corroborating such memories, it doesn't seem at all impossible to confirm.
And no, I'm not taking that back until somebody proves me wrong. So, until you're presented with conclusive evidence of the truth of a thinker's ideas, you'll presume that they are 'frauds' and that their work is 'bullshit'? There's no presumption that they might have something worthwhile on the basis of their fathering almost a new discipline (at the very least a major school) or the continued use of many of the concepts they inaugurated by huge numbers of respected people? Surely you must have some particular grounds for attack - or are we using the 'guilty until proven innocent, and even then if we don't like them' method of jurisprudence pioneered by the Bush administration?
The latter theory has an extra entity but no predictive power. Does it? It seems like in both theories, each subject 'has one entity', which they can use in one way. Which entity are you referring to? The 'potential use' they could make of the memory? Are 'memory' and 'the ability to use memory' both separate objects? (This is why I prefer talking in terms of principles, not entities - I think division into entities is merely formal, not substantive).
Moreover, it would require (given reasonable beliefs about the relationship between memory and the brain) that there be some place in brain where the combined episdoic/procedural memory is encoded that is distinct from the place where each of the two uses is encoded. I don't think I've heard any evidence to that effect. Perhaps. I'm afraid I'm fairly ignorant of reasonable about the relationship between memory and the brain. Last I heard it was a bit of a head-scratcher.
I don't see how one could confirm the existence of anything like "implicit episodic memory." That sounds like Freudian bullshit to me.
Here's the thing, according to this theory, aint no such animal as implicit episodic memory. Implicit memory is broken down into three categories; classical conditioning effects, procedural memory (motor skills, habits 'tacit rules' whatever the fuck that means) and Priming.
so, according to this theory, even a repressed memory is still explicit, there's just some block on its retrieval....
Well, the natural suggestion would seem to be that we confirm it when people 'regain' access to memories that they previously had no access to, such as of events in childhood.
Two reactions: - If it's possible to access memory in an explicit way, it is already explicit memory, since nobody would claim that we are accessing or remembering all our explicit memories all the time. So an implicit memory under this theory would have to be episodic memory that it was in principle impossible to access. - Actually, I take back what I said about it being impossible to confirm implicit episodic memory--we might be able to find it's existence through some route other than personal testimony of internal states. But I don't think there's much evidence to support its existence. What we'd have to show (rigorously) is that something that was functtonally or neurologically comparable to episodic memory existed without somebody's testimony about it.
So, until you're presented with conclusive evidence of the truth of a thinker's ideas, you'll presume that they are 'frauds' and that their work is 'bullshit'?
No. Nothing I've said implies that at all.
There's no presumption that they might have something worthwhile on the basis of their fathering almost a new discipline (at the very least a major school)
To be clear, you're talking about psychoanalysis here, right?
or the continued use of many of the concepts they inaugurated by huge numbers of respected people?
Bandwagon fallacy.
Surely you must have some particular grounds for attack
I do. His methodological bogusness. In his era there was no excuse for not doing psychological work with a rigorously empirical methodology.
Re. implicit declarative memory, yes obviously if the terms are defined that way it would be strikingly hard to detect. Point taken (I have no particular settled opinion here, just testing things out - I do however have a mildly settled opinion regarding psychoanalysis).
Re. psychoanalysis, yes.
Re. 'bandwagon fallacy', that would be a fair claim if the people in question weren't psychologists. Since they are, this is a case where, prior to direct consideration of the issues, the authority of others can give a legitimate presumption of value.
Re. 'methodological bogusness' I think there was, both in his era and in our own, a fine excuse for not doing psychological work with a rigourously empirical method. Rigourously emprical methods work on quantifying things and on dealing with such large numbers of individual cases that each one must be given very little attention. It 'reduces people to mere numbers'. Not that that's an assault of human dignity or anything: it's just unhelpful. We have an excellent, though admittedly fallible, ability to gain understanding of human minds through intensive individual interaction (namely a rationally refined form of basic interpersonal human interaction, which in turn is based on the capacity for insight that derives from the homogeneity of thinker and object and thought). It's stupid to waste this by focusing purely on rigourous empirical methods that while in principle possibly more reliable in practice take many times as long to gain the same level of understanding.
Moreover, empirical methods reveal the empirical features of an object directly (behaviour, neurology, declarations) and the intrinsic features (conscious states) only indirectly and very laboriously. 'Unrigourous methods' can get a form of access to conscious states that, while still indirect insofar as it is another person's consciousness, are direct insofar as two consciousness are of the same type, and of a different type from a physical thing defined behaviourally or neurologically (no, obviously that's not dualism).
that would be a fair claim if the people in question weren't psychologists. Since they are, this is a case where, prior to direct consideration of the issues, the authority of others can give a legitimate presumption of value.
Ok. Your original claim was that it was supported by "huge numbers of respected people." You're right that it's different if you are invoking authorities. However, I can invoke authorities right back and say huge numbers of respected philosophers of science (e.g. Karl Popper) and psychologists (starting with the behaviorists, who were also bad, but that's tangential to the current point) have thought that Freud was pretty sucky.
Anyway, as should be clear, invoking the authority of neo-Freudian psychologists isn't really going to convince me unless you inform me that neo-Freudians aren't making the same mistakes as Freudian ones.
We have an excellent, though admittedly fallible, ability to gain understanding of human minds through intensive individual interaction (namely a rationally refined form of basic interpersonal human interaction, which in turn is based on the capacity for insight that derives from the homogeneity of thinker and object and thought).
I dispute the reliability of this "excellent, admittedly fallible, ability," especially when it's used to form broad theories that are then used to explain our observations of people. It's been shown empirically that even after psychoanalysts have been using things like Rorschach tests and claim to have developed theories that allow them to use them to diagnose patients with this or that, those predictions are at high variance with any statistical correlations. This has been attributed to the psychoanalysts falling prone to confirmation bias, another rigorously empirically tested psychological phenomenon.
The upshot of this is that if we, for example, are sold on a psychodynamic theory of the mind and then sit a lot of people down on a bench and analyse them, then while we may come away convinced that their testimonies are explained by our theory, we could very well be kidding ourselves. We just aren't smart enough to do that kind of thing well without being very, very careful. Like, by using statistics.
Now, that isn't to say that I can't have some kind of empathy with you or some kind of functional folk psychology. But my guess is that a lot of that isn't so much learned as innate, or biological (since most of the theories about causes of autism, which seems to be linked with the absence of that kind of empathy, are neural or biological). But that means that thinking about it abstractly or in a "rationally refined" way isn't likely to capitalize on its successes.
'Unrigourous methods' can get a form of access to conscious states that, while still indirect insofar as it is another person's consciousness, are direct insofar as two consciousness are of the same type, and of a different type from a physical thing defined behaviourally or neurologically (no, obviously that's not dualism).
This sounds like outer space to me. What does being "of the same type" have to do with methodological reliability?
invoking the authority of neo-Freudian psychologists isn't really going to convince me It's not meant to convince you, it's meant to be a reason why, prior to engaging directly with Freud's works, we should give them the benefit of the doubt. This was mainly me trying to draw you into giving a specific problem, rather than saying 'until someone convinces me otherwise'. So that particular discussion is largely moot now.
that means that thinking about it abstractly or in a "rationally refined" way isn't likely to capitalize on its successes. I don't see why. It seems to me that rationally refining it is likely to compensate for its weaknesses.
I dispute the reliability of this "excellent, admittedly fallible, ability,"...those predictions are at high variance with any statistical correlations...What does being "of the same type" have to do with methodological reliability? As you'll notice, I do not deny that the method I'm describing has many unreliabilities. But I think the rigourous empirical methods you prefer are in many ways just as unreliable, precisely because of 'being of the same type'. Because we who theorise are full to the brim with conscious experience we can make theories that are phrased in terms drawn from or definable in terms of the terms of conscious experience. But rigourous empirical methods can only be applied to mathematical-statistical systems. So to use them in psychology, you need to do a conversion from the terms of conscious experience into the terms of mathematics-statistics. Then you can do the mathematical-statistical things involving the theory on the one hand and the data on the other to see how they match up. But the data also require a conversion between mathematical-statistical terms and terms from conscious experience. So you actually have three distinct areas where simplifications, distortions, mistakes, misinterpretations and so forth can creep in.
Moreover, the problems with converting between the two systems are greater the more complex, and hence interesting, the mental things involved. How do you measure bravery in a way that can be subjected to statistical analysis? You have to bring in all sorts of questions and tests and assumptions and stuff. That's why I'm fairly cautious about the 'statistical correlations' you use to rubbish psychoanalysts' predictions. Those correlations were produced by making all kinds of major assumptions and simplifications, which can often be just as, if not more, subject to unconscious bias.
I don't see why. It seems to me that rationally refining it is likely to compensate for its weaknesses.
Well, suppose what we have is an kind of innate capacity for understanding other people. What this has to be (unless we believe in telepathy or some nonsense) is a very reliable mechanism for interpreting our experiences of other people and, perhaps, mapping it onto memories of my own experience.
Now, whatever is helping us with that inference (whatever is helping us to know that such and such actions are indicative of anger) is probably not something that comes immediately to our rational, deductive, theoretical capacity. If we want theoretical knowledge about what helps us with that inference, then we have to abstract away from the results of that empathic capacity.
But once we've done that, we've removed theory from empathy. If I now try to apply the theory to some patient, say, to interpret their behavior, then the prediction is no longer grounded in reliable empathy, but in my ability to shove it into the theoretical framework.
Now, if I continue to build my theory based on these new observations (the one mediated by the psychological theory, and not by empathy), then if I'm not careful I risk turning psychological research into a kind of hermaneutic excercise, where all I need to consider my theory confirmed is the belief that I have successfully interpreted another patient's testimony according to my theory. But that's not empirical confirmation, that's just the textbook case of confirmation bias. It's exactly what David Icke does with the observations he sees about the world: fits them into a framework.
That's what I think happened to Freud. It might have started with some reasonable observations about human behavior ("We have drives, but we also have inhibitions"), but it spins out into what ultimately becomes just a fascinating story. And then Freudian analysis becomes kind of like the Bible Code researchers who use a "methodology" to "discover" secret, prophetic messages from God about the World Trade Center by doing word-finds in scriptural texts.
If you don't think this is the way Freudian theory-building works, I'd be interested to hear how you think one could begin "rationally refining" one's intuitions about other's mental states in a way that wouldn't be improve with more rigorous empirical work.
Because we who theorise are full to the brim with conscious experience we can make theories that are phrased in terms drawn from or definable in terms of the terms of conscious experience.
Yes. Empirical psychologists do this to, where appropriate . This is not true, however, for mental functions that are not necessarily definable exclusively in terms of conscious experience, which is a significant portion of them. Would you agree that for theorizing about these functions this argument does not hold water?
Now, whatever is helping us with that inference (whatever is helping us to know that such and such actions are indicative of anger) is probably not something that comes immediately to our rational, deductive, theoretical capacity. If we want theoretical knowledge about what helps us with that inference, then we have to abstract away from the results of that empathic capacity. I don't accept this. I don't think it's a matter of a set of faculties sitting beside each other and not making eye contact. The 'rational, deductive, theoretical capacity' of 'reason' (I know we've collapsed over this before, I'm planning to post about 'what is reason' in a few days) has, at its heart, such things as synthesis (of current empathetic intuitions with those from other cases, of ideas arising in one area with those in another, etc.) and reflection (the questioning of something previously taken as given so as to move from it to its conditions or detailed nature etc.) and these can be applied to empathetic intuition (I use this for want of a better term) without (at least, without any clear reason that convinces me) 'removing theory from empathy'.
Again, yes, confirmation bias is a issue, and is something that has to be guarded against. I'm not at all recommending that psychology shouldn't use rigourously empirical methods at all - indeed it may well be that Freud's thought would have been improved by their use. But I don't see this as a reason to think that either 1) only rigourously empirical methods should be used, rather than a combination of methods, possibly distributed over different systems or thinkers, or crucially 2) the fact that a thinker's work was one-sided in this respect is grounds for rejecting it outright and calling him a fraud.
This is not true, however, for mental functions that are not necessarily definable exclusively in terms of conscious experience, which is a significant portion of them. Would you agree that for theorizing about these functions this argument does not hold water? Well, a straight answer is tricky, since I have fairly unusual views about the extent of consciousness. I would, though, admit that for theorising about the functions you seem to be talking about, the balance is much more skewed in favour of rigour. So great for them. I think - we may be thinking of different things. I'm thinking of the more 'automatic' functions that usually lie below thresholds of notice. These, though, aren't as interesting, from my point of view, and from the psychoanalytic point of view, as certain other functions.
reason + empathypaulhopeAugust 24 2006, 19:13:44 UTC
The 'rational, deductive, theoretical capacity' of 'reason' (I know we've collapsed over this before, I'm planning to post about 'what is reason' in a few days) has, at its heart, such things as synthesis (of current empathetic intuitions with those from other cases, of ideas arising in one area with those in another, etc.) and reflection (the questioning of something previously taken as given so as to move from it to its conditions or detailed nature etc.) and these can be applied to empathetic intuition (I use this for want of a better term) without (at least, without any clear reason that convinces me) 'removing theory from empathy'.
I don't know what you're saying or, more importantly, why I should accept this account of psychology.
When I'm talking about a 'rational, deductive, theoretical capacity', I'm talking about a specific mental capacity that has been distinguish in cognitive psychology from other capacities in papers such as this one. Specifically, I'm talking about the second, rule-based system. It doesn't seem like that has any direct relation to empathy, so I maintain my position.
Re: reason + empathylukiferAugust 24 2006, 21:57:55 UTC
From what I can see, both these systems seem to be content-neutral. So I apply my secondary processes to intuitive data. I still don't think you've given any actual reason to think that this will result in 'removing' us from the value of the original intuition, any more than applying secondary processes to empirical data somehow destroys the epistemic value that comes from sensory experience. Is this like how boiling vegetables destroys the vitamins?
Also, I think it's funny that the list of people who have had this idea before doesn't include Freud, whose primary process/secondary process idea is precisely this. Apparently he never existed.
Re: reason + empathypaulhopeAugust 25 2006, 14:18:26 UTC
What I said was
"If we want theoretical knowledge about what helps us with that inference, then we have to abstract away from the results of that empathic capacity."
I.e., there is an extra step of inference from "intuitive data" into some kind of theoretical representation on which the secondary process can work (and which, ultimately, can be communicated, etc.) That abstraction into theory removes it from the kind of fact-checking we might get from intuition and leaves it prone to confirmation bias, etc.
This isn't so bad in our day-to-day life--there's still some "epistemic value" there, I guess. But as a social science, it's a disaster. Without rigor, it turns into superstition and hermeneutics.
Re: reason + empathylukiferAugust 29 2006, 14:02:52 UTC
there is an extra step...[that] removes it from the kind of fact-checking we might get from intuition I suppose that's true to a degree. But not too big.
But as a social science, it's a disaster. Without rigor, it turns into superstition and hermeneutics. See, I think what this is likely to come down to is us both recognising the problems the other is talking about but saying 'yeah but your problems are BIGGER'.
Also, I'm curious what you mean by 'hermeneutics', since you seem very keen on it when it's Bayesian.
Well, I guess, but then I'm not too sure about any distinction between memory and knowledge. If I learn that south american monkeys have a different kind of nose from old world monkeys, and I remember that, is that memory or knowledge? Similarly, if I learn to ride a bike, is that procedural memory or procedural knowledge? I'm honestly asking, is there some clear distinction that's drawn? 'Cos it seems to me that 'memory' can be stretched to any information stored in the mind that's derived from past experience, which is almost all of it. And if knowledge counts, then belief would have to count, insofar as they differ in terms of how reality is. And I think beliefs and attitudes blur into one another ('I hate him' blurs into 'he is a bastard').
And Freud was a fraud.
You take that back!
if your hypothesis (that most memories have some of each), it would be much harder to explain these cases.
Well fine, I guess - I'm not (yet) an empirical neurophysiologist, so I can't say much here, though for what it's worth I don't think there's too much of a incompatibility here: if any given memory can have two aspects when it's used in two ways, then it would probably be used 'by' two different loci. I suppose though it then becomes really a semantic issue (in the bad sense) whether we say that there are 'two memories' or one memory processed in two different ways.
Sorry to be slow to get back to you on your metaphysics opus
No problem. Wrists are demanding mistresses.
Reply
ok, here's where psychologists would deploy the semantic v. episodic memory distinction, sort of a weak-ass attempt to separate abstractions and propositions from the garbled mass of associations which can't be distilled into simple sentences/propositions or beliefs, but still kick around in our head.
zB. -
your experience when learning this fact about monkeys, filled with all sorts of extraneous information - the tone of voice that the speaker related the information to you, the brightness of your environment, the feeling of your chafing boxer-briefs- is all supposedly, imprinted into our "episodic memory" and cannot be recalled as a collection of propositions, just one baffling chaotic experience. After the fact, we then pull atomic or simple propositions out of this mess. "my boxers were chafing", "south american monkeys have a different type of nose." and so on. interestingly, in this sort of analysis, there's no ability to generalize about experience by just calling the episodic memory. the only things we can report later on and hence (?) "know" are stored in semantic memory. I envision it with the help of the analogy of a transparency used as a visual aid. We have a jumbled image below (our episodic memory) and then above we have a transparent overlay, with little boxes containing the propositions contained in our semantic memory. There should be some mechanism that leaves little nodes or anchors in the episode that connect to the propositions.
psychologists apparently don't distinguish between memory and knowledge per se. (or at least those who have informed what i've read in my low-level psych courses, i'm sure there ARE psychologists who maintain this distinction, i'd wager those of the philosopher persuasion DO) Anyway, it seems like 'knowledge', which is by nature abstracted (?), can be seen as the content of our 'semantic' store
i think, however, conceptions of knowledge and the dividing line between bodily functions/abilities and knowledge that one can articulate is an interesting topic for discussion. for example, swamis or yogis in india consider your autonomic processes 'knowledge' which most western philosophers would balk at, like woah.
anyway, my roommates are drunk and boisterous and this has been incoherent enough already. this house is not conducive to thought on ANY level.
Reply
I stand corrected by my buddy.
And, after reading both your comments, and remembering some stuff I should have remembered earlier about different systems of reasoning, I'm going to change my position somewhat:
I think it would be appropriate to blur the line on procedural memory/knowledge. And I think that it is implicit in that we can't consciously access it declare it.
And I think that it is a different kind of memory (not just a different use of the same memory) than episodic and semantic memories.
I don't see how one could confirm the existence of anything like "implicit episodic memory." That sounds like Freudian bullshit to me. And no, I'm not taking that back until somebody proves me wrong.
For semantic memory, I remember now that there is a supported hypothesis for two systems of reasoning, one implicit and one explicit, which would work using something like semantic knowledge/memory. If I remember right, it was an open question whether or not it was the same knowledge being used in two ways, or two different kinds of knowledge, although I think the prof who was teaching me about it believed strongly that it was the latter.
-----
I don't know where that puts me relative to your intuitions. Really, I ought to look all this stuff up again.
But I disagree with the claim that the "two use" and "two memories" theories are identical except for semantics. Consider the episodic/procedural case. Suppose two subjects are trained in a task, and one, subject A, can remember the training but is not any better at performing the task for it, and the other, subject B, can't remember the training but is good at the task (this has happened.)
Now, either we say:
(1) A does not have the procedural memory but has the episodic memory, and the opposite is the case for B.
or
(2) A still has the combine procedural/episodic memory of the training experience, but has simply lost the procedural use of that memory. The evidence for the original memory (which could potentially have both uses) is the observable episodic memory.
The latter theory has an extra entity but no predictive power. So if only on the grounds of parsimony, it should be rejected. Moreover, it would require (given reasonable beliefs about the relationship between memory and the brain) that there be some place in brain where the combined episdoic/procedural memory is encoded that is distinct from the place where each of the two uses is encoded. I don't think I've heard any evidence to that effect.
Reply
Word. that's tricky business, a distinct encoding area for the memory as opposed to the use. i dont think i buy it either.
but:
here's what im thinking: (colon inside a colon, weird)
episodic memory isn't productive, in the sense that (maybe due to its unruliness) episodic memories cant directly influence behavior or self-report. they have to be translated into component parts, pared down, specified to particular tasks, etc. what i refered to as 'semantic memory' on the other hand, is production ready, its the actor's formula for action, belief, thought. i dunno, it seems as though episodic memory would have to be more primal, and that as soon as you begin to really analyze it, you've started shifting it over into semantic memory, using abstract characteristics/labels/whatever to describe it.
which birngs about the question of form.
my thoughts sort of presuppose that there is NOT some uniformity in encoding between the two memories, but rather a function (or suite of functions) that maps parts of episodic memory to semantic. No matter how hard i try, i cannot bring myself to buy into mentalese wholeheartedly.
maybe mentalese is used for cross-compartmental correspondence, but i think it would only be as a last resort. things wouldnt be translated into mentalese unless they absolutely HAVE to be.
so, episodic memory is a MESS, because its got tons of different types of inputs and coding and what not. (at the very least info from the 5 senses, and then some emotional information, at the most basic level), but it neednt be encoded into a single form or language.
but then again we're back to a problem - this procedural knowledge that one CAN'T report on, is it in mentalese? Suppose we posit that 'knowledge', which i think is just a subset of memory, is all in mentalese, as to facilitate production and transmission of information, WHERE THE HELL does the explicit/implicit demarcation fall? if 'physical' knowledge, like the autonomic system's functions, is also stored in mentalese - why CAN"T i report on it?
Reply
Well, the natural suggestion would seem to be that we confirm it when people 'regain' access to memories that they previously had no access to, such as of events in childhood. That's not to deny that there aren't cases where such memories could be mere fantasies, but insofar as there is the possibility of corroborating such memories, it doesn't seem at all impossible to confirm.
And no, I'm not taking that back until somebody proves me wrong.
So, until you're presented with conclusive evidence of the truth of a thinker's ideas, you'll presume that they are 'frauds' and that their work is 'bullshit'? There's no presumption that they might have something worthwhile on the basis of their fathering almost a new discipline (at the very least a major school) or the continued use of many of the concepts they inaugurated by huge numbers of respected people? Surely you must have some particular grounds for attack - or are we using the 'guilty until proven innocent, and even then if we don't like them' method of jurisprudence pioneered by the Bush administration?
The latter theory has an extra entity but no predictive power.
Does it? It seems like in both theories, each subject 'has one entity', which they can use in one way. Which entity are you referring to? The 'potential use' they could make of the memory? Are 'memory' and 'the ability to use memory' both separate objects? (This is why I prefer talking in terms of principles, not entities - I think division into entities is merely formal, not substantive).
Moreover, it would require (given reasonable beliefs about the relationship between memory and the brain) that there be some place in brain where the combined episdoic/procedural memory is encoded that is distinct from the place where each of the two uses is encoded. I don't think I've heard any evidence to that effect.
Perhaps. I'm afraid I'm fairly ignorant of reasonable about the relationship between memory and the brain. Last I heard it was a bit of a head-scratcher.
Reply
Here's the thing, according to this theory, aint no such animal as implicit episodic memory. Implicit memory is broken down into three categories; classical conditioning effects, procedural memory (motor skills, habits 'tacit rules' whatever the fuck that means) and Priming.
so, according to this theory, even a repressed memory is still explicit, there's just some block on its retrieval....
Reply
Two reactions:
- If it's possible to access memory in an explicit way, it is already explicit memory, since nobody would claim that we are accessing or remembering all our explicit memories all the time. So an implicit memory under this theory would have to be episodic memory that it was in principle impossible to access.
- Actually, I take back what I said about it being impossible to confirm implicit episodic memory--we might be able to find it's existence through some route other than personal testimony of internal states. But I don't think there's much evidence to support its existence. What we'd have to show (rigorously) is that something that was functtonally or neurologically comparable to episodic memory existed without somebody's testimony about it.
So, until you're presented with conclusive evidence of the truth of a thinker's ideas, you'll presume that they are 'frauds' and that their work is 'bullshit'?
No. Nothing I've said implies that at all.
There's no presumption that they might have something worthwhile on the basis of their fathering almost a new discipline (at the very least a major school)
To be clear, you're talking about psychoanalysis here, right?
or the continued use of many of the concepts they inaugurated by huge numbers of respected people?
Bandwagon fallacy.
Surely you must have some particular grounds for attack
I do. His methodological bogusness. In his era there was no excuse for not doing psychological work with a rigorously empirical methodology.
Reply
Re. psychoanalysis, yes.
Re. 'bandwagon fallacy', that would be a fair claim if the people in question weren't psychologists. Since they are, this is a case where, prior to direct consideration of the issues, the authority of others can give a legitimate presumption of value.
Re. 'methodological bogusness' I think there was, both in his era and in our own, a fine excuse for not doing psychological work with a rigourously empirical method. Rigourously emprical methods work on quantifying things and on dealing with such large numbers of individual cases that each one must be given very little attention. It 'reduces people to mere numbers'. Not that that's an assault of human dignity or anything: it's just unhelpful. We have an excellent, though admittedly fallible, ability to gain understanding of human minds through intensive individual interaction (namely a rationally refined form of basic interpersonal human interaction, which in turn is based on the capacity for insight that derives from the homogeneity of thinker and object and thought). It's stupid to waste this by focusing purely on rigourous empirical methods that while in principle possibly more reliable in practice take many times as long to gain the same level of understanding.
Moreover, empirical methods reveal the empirical features of an object directly (behaviour, neurology, declarations) and the intrinsic features (conscious states) only indirectly and very laboriously. 'Unrigourous methods' can get a form of access to conscious states that, while still indirect insofar as it is another person's consciousness, are direct insofar as two consciousness are of the same type, and of a different type from a physical thing defined behaviourally or neurologically (no, obviously that's not dualism).
Reply
Ok. Your original claim was that it was supported by "huge numbers of respected people." You're right that it's different if you are invoking authorities. However, I can invoke authorities right back and say huge numbers of respected philosophers of science (e.g. Karl Popper) and psychologists (starting with the behaviorists, who were also bad, but that's tangential to the current point) have thought that Freud was pretty sucky.
Anyway, as should be clear, invoking the authority of neo-Freudian psychologists isn't really going to convince me unless you inform me that neo-Freudians aren't making the same mistakes as Freudian ones.
We have an excellent, though admittedly fallible, ability to gain understanding of human minds through intensive individual interaction (namely a rationally refined form of basic interpersonal human interaction, which in turn is based on the capacity for insight that derives from the homogeneity of thinker and object and thought).
I dispute the reliability of this "excellent, admittedly fallible, ability," especially when it's used to form broad theories that are then used to explain our observations of people. It's been shown empirically that even after psychoanalysts have been using things like Rorschach tests and claim to have developed theories that allow them to use them to diagnose patients with this or that, those predictions are at high variance with any statistical correlations. This has been attributed to the psychoanalysts falling prone to confirmation bias, another rigorously empirically tested psychological phenomenon.
The upshot of this is that if we, for example, are sold on a psychodynamic theory of the mind and then sit a lot of people down on a bench and analyse them, then while we may come away convinced that their testimonies are explained by our theory, we could very well be kidding ourselves. We just aren't smart enough to do that kind of thing well without being very, very careful. Like, by using statistics.
Now, that isn't to say that I can't have some kind of empathy with you or some kind of functional folk psychology. But my guess is that a lot of that isn't so much learned as innate, or biological (since most of the theories about causes of autism, which seems to be linked with the absence of that kind of empathy, are neural or biological). But that means that thinking about it abstractly or in a "rationally refined" way isn't likely to capitalize on its successes.
'Unrigourous methods' can get a form of access to conscious states that, while still indirect insofar as it is another person's consciousness, are direct insofar as two consciousness are of the same type, and of a different type from a physical thing defined behaviourally or neurologically (no, obviously that's not dualism).
This sounds like outer space to me. What does being "of the same type" have to do with methodological reliability?
Reply
It's not meant to convince you, it's meant to be a reason why, prior to engaging directly with Freud's works, we should give them the benefit of the doubt. This was mainly me trying to draw you into giving a specific problem, rather than saying 'until someone convinces me otherwise'. So that particular discussion is largely moot now.
that means that thinking about it abstractly or in a "rationally refined" way isn't likely to capitalize on its successes.
I don't see why. It seems to me that rationally refining it is likely to compensate for its weaknesses.
I dispute the reliability of this "excellent, admittedly fallible, ability,"...those predictions are at high variance with any statistical correlations...What does being "of the same type" have to do with methodological reliability?
As you'll notice, I do not deny that the method I'm describing has many unreliabilities. But I think the rigourous empirical methods you prefer are in many ways just as unreliable, precisely because of 'being of the same type'. Because we who theorise are full to the brim with conscious experience
we can make theories that are phrased in terms drawn from or definable in terms of the terms of conscious experience. But rigourous empirical methods can only be applied to mathematical-statistical systems. So to use them in psychology, you need to do a conversion from the terms of conscious experience into the terms of mathematics-statistics. Then you can do the mathematical-statistical things involving the theory on the one hand and the data on the other to see how they match up. But the data also require a conversion between mathematical-statistical terms and terms from conscious experience. So you actually have three distinct areas where simplifications, distortions, mistakes, misinterpretations and so forth can creep in.
Moreover, the problems with converting between the two systems are greater the more complex, and hence interesting, the mental things involved. How do you measure bravery in a way that can be subjected to statistical analysis? You have to bring in all sorts of questions and tests and assumptions and stuff. That's why I'm fairly cautious about the 'statistical correlations' you use to rubbish psychoanalysts' predictions. Those correlations were produced by making all kinds of major assumptions and simplifications, which can often be just as, if not more, subject to unconscious bias.
Reply
Well, suppose what we have is an kind of innate capacity for understanding other people. What this has to be (unless we believe in telepathy or some nonsense) is a very reliable mechanism for interpreting our experiences of other people and, perhaps, mapping it onto memories of my own experience.
Now, whatever is helping us with that inference (whatever is helping us to know that such and such actions are indicative of anger) is probably not something that comes immediately to our rational, deductive, theoretical capacity. If we want theoretical knowledge about what helps us with that inference, then we have to abstract away from the results of that empathic capacity.
But once we've done that, we've removed theory from empathy. If I now try to apply the theory to some patient, say, to interpret their behavior, then the prediction is no longer grounded in reliable empathy, but in my ability to shove it into the theoretical framework.
Now, if I continue to build my theory based on these new observations (the one mediated by the psychological theory, and not by empathy), then if I'm not careful I risk turning psychological research into a kind of hermaneutic excercise, where all I need to consider my theory confirmed is the belief that I have successfully interpreted another patient's testimony according to my theory. But that's not empirical confirmation, that's just the textbook case of confirmation bias. It's exactly what David Icke does with the observations he sees about the world: fits them into a framework.
That's what I think happened to Freud. It might have started with some reasonable observations about human behavior ("We have drives, but we also have inhibitions"), but it spins out into what ultimately becomes just a fascinating story. And then Freudian analysis becomes kind of like the Bible Code researchers who use a "methodology" to "discover" secret, prophetic messages from God about the World Trade Center by doing word-finds in scriptural texts.
If you don't think this is the way Freudian theory-building works, I'd be interested to hear how you think one could begin "rationally refining" one's intuitions about other's mental states in a way that wouldn't be improve with more rigorous empirical work.
Because we who theorise are full to the brim with conscious experience
we can make theories that are phrased in terms drawn from or definable in terms of the terms of conscious experience.
Yes. Empirical psychologists do this to, where appropriate . This is not true, however, for mental functions that are not necessarily definable exclusively in terms of conscious experience, which is a significant portion of them. Would you agree that for theorizing about these functions this argument does not hold water?
Reply
I don't accept this. I don't think it's a matter of a set of faculties sitting beside each other and not making eye contact. The 'rational, deductive, theoretical capacity' of 'reason' (I know we've collapsed over this before, I'm planning to post about 'what is reason' in a few days) has, at its heart, such things as synthesis (of current empathetic intuitions with those from other cases, of ideas arising in one area with those in another, etc.) and reflection (the questioning of something previously taken as given so as to move from it to its conditions or detailed nature etc.) and these can be applied to empathetic intuition (I use this for want of a better term) without (at least, without any clear reason that convinces me) 'removing theory from empathy'.
Again, yes, confirmation bias is a issue, and is something that has to be guarded against. I'm not at all recommending that psychology shouldn't use rigourously empirical methods at all - indeed it may well be that Freud's thought would have been improved by their use. But I don't see this as a reason to think that either 1) only rigourously empirical methods should be used, rather than a combination of methods, possibly distributed over different systems or thinkers, or crucially 2) the fact that a thinker's work was one-sided in this respect is grounds for rejecting it outright and calling him a fraud.
This is not true, however, for mental functions that are not necessarily definable exclusively in terms of conscious experience, which is a significant portion of them. Would you agree that for theorizing about these functions this argument does not hold water?
Well, a straight answer is tricky, since I have fairly unusual views about the extent of consciousness. I would, though, admit that for theorising about the functions you seem to be talking about, the balance is much more skewed in favour of rigour. So great for them. I think - we may be thinking of different things. I'm thinking of the more 'automatic' functions that usually lie below thresholds of notice. These, though, aren't as interesting, from my point of view, and from the psychoanalytic point of view, as certain other functions.
Reply
I don't know what you're saying or, more importantly, why I should accept this account of psychology.
When I'm talking about a 'rational, deductive, theoretical capacity', I'm talking about a specific mental capacity that has been distinguish in cognitive psychology from other capacities in papers such as this one. Specifically, I'm talking about the second, rule-based system. It doesn't seem like that has any direct relation to empathy, so I maintain my position.
Reply
Also, I think it's funny that the list of people who have had this idea before doesn't include Freud, whose primary process/secondary process idea is precisely this. Apparently he never existed.
Reply
"If we want theoretical knowledge about what helps us with that inference, then we have to abstract away from the results of that empathic capacity."
I.e., there is an extra step of inference from "intuitive data" into some kind of theoretical representation on which the secondary process can work (and which, ultimately, can be communicated, etc.) That abstraction into theory removes it from the kind of fact-checking we might get from intuition and leaves it prone to confirmation bias, etc.
This isn't so bad in our day-to-day life--there's still some "epistemic value" there, I guess. But as a social science, it's a disaster. Without rigor, it turns into superstition and hermeneutics.
Reply
I suppose that's true to a degree. But not too big.
But as a social science, it's a disaster. Without rigor, it turns into superstition and hermeneutics.
See, I think what this is likely to come down to is us both recognising the problems the other is talking about but saying 'yeah but your problems are BIGGER'.
Also, I'm curious what you mean by 'hermeneutics', since you seem very keen on it when it's Bayesian.
Reply
Leave a comment