Over at Ktismatics John has an interesting post up making the case that we have direct access to our minds. John writes:
In my post I summarized some of the empirical evidence supporting the contention that much, if not most, human cognition takes place outside of conscious awareness. However, I decidedly did not propose that all of cognition is unconscious. We consciously attend to things, reason, solve problems, assemble stored memories, plan, evaluate information. And we’re self-reflexive about it: we are consciously aware that we’re reasoning, problem-solving, etc.
Doesn’t this mean that we have direct access to our own minds, at least to some extent? I’d say yes. If’ I’m aware that I’m solving a problem, and if both my awareness and my problem-solving are mental processes, then my mind has direct access to some of its own activities. If we’re consciously aware of the activities and outputs of our own consciousness, then that’s not just direct self-relation but also direct self-awareness of the self-relation. Consciousness is an emergent property of brain activity; unless we believe in the soul or some form of panpsychism there is no source of consciousness other than brain activity. So consciousness has to be in direct relation with the unconscious brain activity that generates it — doesn’t it? — even if that direct relation doesn’t take the form of conscious awareness of brain function. My hand is in direct connection with itself, even if it can’t hold itself in its grip. A bridge is in direct connection with itself, even if it can’s support itself on itself.
Yes and no. The claim that we do not have direct access to our minds is not the claim that our minds are not dependent on our brains, but rather that our introspective accounts of our mental activities are unreliable guides to how these mental activities function. In other words, the thesis that we can have knowledge of how our mental states function through self-reflexive conscious awareness of our internal states is here challenged.
read on!
To understand why this observation is philosophically important it is necessary to understand how claims about conscious awareness have functioned in the history of philosophy. In his second meditation, Descartes writes:
8. But what, then, am I ? A thinking thing, it has been said. But what is a thinking thing? It is a thing that doubts, understands, [conceives], affirms, denies, wills, refuses; that imagines also, and perceives. [ L][ F]
9. Assuredly it is not little, if all these properties belong to my nature. But why should they not belong to it ? Am I not that very being who now doubts of almost everything; who, for all that, understands and conceives certain things; who affirms one alone as true, and denies the others; who desires to know more of them, and does not wish to be deceived; who imagines many things, sometimes even despite his will; and is likewise percipient of many, as if through the medium of the senses. Is there nothing of all this as true as that I am, even although I should be always dreaming, and although he who gave me being employed all his ingenuity to deceive me ? Is there also any one of these attributes that can be properly distinguished from my thought, or that can be said to be separate from myself ? For it is of itself so evident that it is I who doubt, I who understand, and I who desire, that it is here unnecessary to add anything by way of rendering it more clear. And I am as certainly the same being who imagines; for although it may be (as I before supposed) that nothing I imagine is true, still the power of imagination does not cease really to exist in me and to form part of my thought. In fine, I am the same being who perceives, that is, who apprehends certain objects as by the organs of sense, since, in truth, I see light, hear a noise, and feel heat. But it will be said that these presentations are false, and that I am dreaming. Let it be so. At all events it is certain that I seem to see light, hear a noise, and feel heat; this cannot be false, and this is what in me is properly called perceiving (sentire), which is nothing else than thinking.
Descartes’ Meditations, especially the first three meditations, is essential reading on these issues. In the passage above Descartes argues that thought provides a ground of absolute certainty or incorrigibility, and is therefore the foundation or ground upon which all knowledge must be based. If, according to Descartes, thought has this special characteristic then this is because it is absolutely present to itself without any accompanying absence.
To understand the epistemological significance of this self-presence, it is necessary to understand just how re-presentation contains the possibility of error woven within it. Rather than beginning with thorny questions of how mind represents the world, it would perhaps be more illuminating to look at more common examples of representation from the worlds of law and politics. In law and politics a representative is a third party that comes between two people or groups of people, mediating their relation to one another. For example, a diplomat represents one nation to another nation. Neither nation relates directly to the other nation, but rather each nation only relates directly to the representative. This entails that any knowledge one of the nations has about the other nature is always received second-hand through the representative.
Here the relationship between representation and absence becomes evident. Insofar as the relationship between the two nations is mediated by a third party, insofar as their relationship to one another is indirect, the two parties are absent to one another and must rely on the reliability and honesty of the representative or diplomat for their access to each other. Consequently, mediated relations between parties always contain an element of doubt or uncertainty. The case is no different with representational theories of knowledge. The representation is a third term that comes between the knowing subject and the world to be known. What our minds relate to is not the object but our representations of objects. In this respect, representations have the structure of signs. As C.S. Peirce famously defined them, a sign is something that stands for something in some respect or capacity. In a clever twist, Umberto Eco, in A Theory of Semiotics, defines a sign as anything that can be used to deceive or tell a lie. Representations are entities that purport to stand for something else. As such, signs necessarily embody an absence insofar as what we directly relate to is not the world but the representation.
If signs and representations necessarily contain a degree of uncertainty or doubt, then this is because they embody absence within themselves, implying a present that is not, in fact, present. This point can be illustrated with reference to the final scenes of Kubric’s The Shining [around the 3:13 mark]:
Running through the topiary in the midst of a blizzard, the boy escapes the clutches of his crazed father through the skillful use of signs. The signs in question are his footprints in the snow. These footprints stand for the path the boy has taken. The father uses these signs to follow the boy’s trail. The boy, possessing a basic understanding of signs, steps back in his own footprints and then runs off in another direction, covering his tracks, so that his father will be led astray. Here the signs deceive his father and the boy fortunately escapes.
Here then we encounter ground zero of the question of our knowledge. Given that our signs or representations are the “presence of an absence” (i.e., the presence of something we are not directly related to), how can we determine the difference between those signs or representations that are veridical or accurate representations of the world and those that are not? If we are only directly related to our representations and not the world, then the question appears to be insoluble. Descartes’ sought an absolute foundation for knowledge that was beyond all doubt. Clearly our representations will not deliver us this foundation because they embody an absence that perpetually leaves room for doubt. What is needed is an absolute presence that would establish certainty. If thought, according to Descartes, provides such a foundation, then this is because my relationship to my thought is characterized by complete presence or immediacy (i.e., the absence of mediation) such that while I can doubt what my thought represents or stands for, I cannot doubt the simple having of this thought itself. While I can doubt whether or not I’m actually typing these words or am merely dreaming that I’m typing these words, I cannot doubt, according to Descartes, that I seem to be having this experience.
Thought, for Descartes, is thus unassailable or beyond all doubt because it embodies no absence or is characterized by immediate self-evidence and certainty. Initially Descartes’ gain here appears to be slight as we’re still in doubt about the existence of the world, other minds, our own bodies, our memories, and so on. However, Descartes does believe that he has opened up the domain of thought as a reliable field of investigation that cannot be doubted. The issue will now be one of determining whether or not we can, based on thought alone, establish the transcendence of the world in a way that is absolutely certain and this will require an investigation of the contents of our thought.
I will not here rehearse in detail the remainder of Descartes’ argument. Readers will find it in the third meditation. According to Descartes, the first thing of which we can be certain beyond ourselves is the existence of God. Since God is perfect he is therefore good. Insofar as God is good he cannot be a deceiver. Therefore, thinks Descartes, we can trust certain representations we possess as being accurate representations of the world.
Having briefly taken a detour through Descartes, it is now possible to re-situate the question of direct access with respect to our minds. Do we have, as Descartes suggested, direct access to our thought as Descartes suggested? Subsequent history and investigation has not been very supportive of Descartes’ hypothesis. In particular, neurology and psychoanalysis have both shed a great deal of doubt on Descartes’ equation of thought with awareness. John wrote:
We consciously attend to things, reason, solve problems, assemble stored memories, plan, evaluate information. And we’re self-reflexive about it: we are consciously aware that we’re reasoning, problem-solving, etc.
Is the suggestion that we are conscious or aware of doing these things really as self-evident as John suggests? I think not. Take our moral reasoning or problem-solving. In his recent book How We Decide (which is a terrific read, even if I disagree with it on many points), neurologist Jonah Lehrer recounts fMRI experiments with subjects engaged in moral deliberation. What these scans show is that moral decision making takes place primarily and in the first place in the amygdala, with the frontal cortex only subsequently becoming active after the decision has already been made. Moreover, those subjects that have significant damage or abnormality to their amygdala, transforming them into perfect Kantian calculative subjects, are often severely impaired where moral decision making is concerned. There is a strong connection here between sociopathy and such abnormalities among such subjects.
Now the amygdala is a region of the brain strongly connected to emotions and memory, while the frontal cortex is related to operations of higher reasoning. The significance of these observations with regard to the thesis that equates thought with awareness should be evident. Subjects who are asked why they wouldn’t steal candy at a grocery store will appeal to reasons such as potential consequences of their actions, the violation of some abstract moral principle like the categorical imperative or one of the Ten Commandments, etc. The fMRI scans reveal something different. Far from being a deliberative problem-solving process, the decision is already made at the level of unconscious emotional reactions. In other words, the higher-order problem solving activity comes ex post facto. The subject experiences these higher order reasoning processes as what led him to the decision, but the decision was already made behind his back at an unconscious level.
The conclusion to be drawn from this example– and examples of such findings could be multiplied infinitely –is that thought and awareness are not the same thing. There is a profound disadequation between how we experience ourselves (the domain of conscious awareness) and what produces these experiences (the domain of thought). Our self-reports based on conscious introspection are tremendously unreliable as guides as to why we do things. Consequently, when John writes:
If we’re consciously aware of the activities and outputs of our own consciousness, then that’s not just direct self-relation but also direct self-awareness of the self-relation. Consciousness is an emergent property of brain activity; unless we believe in the soul or some form of panpsychism there is no source of consciousness other than brain activity.
I cannot help but feel he’s missed the basic point of these discussions. This is clear from his references to panpsychism and the soul which are irrelevant to the issue under discussion. When a philosopher claims that we do not have direct access to our minds he is not claiming that the brain doesn’t produce the mind or that mind is not an emergent property of brain. One can adopt (as I do) the thesis that mind is an emergent property of brain and that we do not have direct access to our minds without being a panpsychist. The point here is epistemological, not ontological. What is at issue is whether conscious thought is a privileged domain of knowledge that can function as an absolutely secure foundation for all other knowledge as thinkers like Descartes, Hume (with his impressions or sensations), Husserl, and many others besides have argued. What researches in fields such as psychoanalysis and neurology have revealed is that far from being such a privileged and foundational domain, our relation to our own thought is every bit as deceptive and prone to error as our relationship to “external” objects in the world. As a consequence, there is no reason to grant conscious awareness a privileged and foundational role in philosophical thought. Our introspective reports as to why we think as we do are representational with respect to ourselves no less than with respect to objects (thereby embodying all the doubt we encountered in the relationship between representations and objects represented), and our introspective accounts of how thought actually functions are very likely to be extremely misleading and outright false. Yet if this point is granted, a number of assumptions, central axioms, and problematics governing contemporary philosophy are transformed significantly.
November 30, 2009 at 8:20 pm
Thanks for the link and interaction, Levi. I was just about to take an after-lunch nap when I saw your post. I tried napping, but my mind was working away on possible replies and rejoinders. Realizing that I was thinking too much to doze off, I tried distracting myself, but to no avail. So here I am.
In general I’d say that my claims are far more modest than those you ascribe to me. You cite the beginning of my post in which I affirm that “much, if not most, human cognition takes place outside of conscious awareness.” However, and I hope not controversially, some cognition is conscious. Reasoning, problem-solving, planning, and so on are the sorts of cognitive activities that take place at least partly in consciousness. Still, in order to perform these conscious tasks we need to retrieve memories, scripts, schemata, and sequences that have been compiled and stored in the brain beneath the threshold of conscious awareness. As I wrote later in the post, “What I definitely don’t have direct access to are the unconscious workings of the mind, which mostly have to do with the neural structures and synaptic firings from which my conscious thoughts arise.” In other words, I can know that I’m thinking without knowing how I’m thinking at the synaptic level.
The point is, if I know that I am thinking, then I have at least some direct access to the workings of my mind. To insist that only neurological knowledge of cognition counts as self-awareness is, I think, to take an overly-reductionistic approach to cognition. I’m sure you’d agree that thoughts emerge from but aren’t the same thing as neural firing patterns. I can manipulate thoughts consciously, and be consciously aware that I’m doing so, without having direct knowledge of how my brain is assembling those thoughts.
As I said, it’s a rather modest claim I’m making. I certainly don’t propose that our cognitive self-knowledge is complete, or fully reliable, or philosophically foundational. I also don’t address representational theories of knowledge. I’m only saying that, while limited in important and demonstrable ways, cognitive self-awareness does happen.
In putting together my post I was attempting to respond to what I regarded as unnecessarily excessive claims that you had previously made here at LS. I don’t have the specific posts listed, but I did write down a couple of your quotes:
“if the last 300 years of philosophy have shown us anything, it has shown us that we do not have any direct or immanent or immediate access to our own minds.”
“No credible source has been able to establish direct self-relation since the advent of neurology and the unconscious”
Again, I regarded your assertions as perhaps overly reductionistic in demanding that we have direct intuition into how we think at the neurological level. I didn’t think this kind of argument fit well with the general tenor of your ontology, but maybe I’m wrong. If, e.g., I’m thinking about the names of the people who lived in my childhood neighborhood, and if I’m aware that I’m thinking about it, that should be enough to demonstrate my mind’s direct access to itself. The other possibility is to claim that there’s a self-awareness entity in my head that’s different from the memory retrieval entity, which I presume you don’t have in mind either.
Maybe you could elaborate on why it’s important to your theory that minds do not have direct access to themselves. Or perhaps you could explain why my being aware that I am thinking doesn’t count as direct cognitive self-awareness.
November 30, 2009 at 8:50 pm
Thanks John. I don’t disagree with much of what you say here. I did sense that you were responding to me. Given that you were responding to me it is important that the philosophical issues and importance of these points be properly understood. With respect to these issues, there is nothing “excessive” in claims I have made such as you suggest. The privilege of consciousness and immediate access to consciousness has been one of the lynchpins of foundationalist epistemologies, prohibiting all sorts of philosophical moves on the grounds that they cannot be demonstrated based on the direct and unassailable evidences of consciousness (Husserl). The point is not that we have no access to our minds, but that that access is very limited and often prone to error. As a consequence conscious awareness cannot serve as the foundation that many proponents of phenomenology and idealism would like to make of it. That is all.
It is important to distinguish between criticism of a claim and advocacy of a claim. When I argue that we do not have direct access to the contents of our own mind I am taking the central claims of other philosophical positions seriously and at their word and showing why they fail to hold up under scrutiny. There is nothing reductivist in this argument. It is the claims that folks like Descartes and Husserl actually make. I am showing why these claims to absolute foundations do not hold and thereby deflating the claims these positions make to rigor and apodicticity. As for your example about names, I think my example about moral reasoning already puts into question any direct access we might have to the reasons we have the thoughts we do. That’s what’s important here.
Why is all this important? To understand why this is important background in the history of philosophy and the axiomatic assumptions of philosophy is necessary. I believe I explained a good deal of this in my post. If the direct access argument holds, then philosophically it would be established that anything that does not base itself on the certitude of consciousness is dogmatic. Insofar as one form of philosophy argues that we must get rid of all undemonstrated presuppositions, the direct access argument limits us more or less to phenomenology as the only legitimate form of philosophy, rejecting any claims that can’t be demonstrated in terms of direct and immediate awareness to the dogmatic. If demonstrating that our access to our own thought is not incorrigible is so important, then this is because doing so significantly opens the domain of what we can investigate philosophically be ousting consciousness from its privileged place as an epistemic foundation.
In my view your observation that we are aware that we are thinking is sufficient criteria is way too loose and fuzzy to be of any philosophical use. What exactly is it you’re aware of when you’re thinking? Do you actually know the reasons you’re thinking what you’re thinking, how those thoughts came to be, etc? Sure, I am aware that I find it morally repugnant to push a man off a train platform to stop a train. What I am unaware of is the cognitive process that leads me to this conclusion. What fields like neurology and psychoanalysis reveal is that the processes by which this conscious state produce are often wildly at odds with the reasons we would consciously give for why we have this thought. As such, they undermine the thesis that we have immediate access to our thought.
November 30, 2009 at 10:21 pm
A friend of mine once said “without the exponential function we can’t move forward”. Herein lies the problem.
Descartes gave us a precious gift, that of identity – a self-loop which confirms the presented content while being performed, which is the content – but the identity isn’t enough. We need exponentials.
Levi’s deceptive god is even closer than Descartes. With the brain as the mediator of conscious events there is no hope for reliable confirmations of coherence beyond the identity – no exponentials. The good news is that sensual perceptions live on equal footing. All of them are unreliable but the deceptive god exists in symbiosis with the conscious mind and will do whatever possible to keep the error small, the illusion perfect and the thinking accurate. There is nothing in the universe which is as helpful as the deceptive god and bad luck if it fails.
The moral of the story is that we cannot take the identity as hook for performing better than the whole system. We cannot trick the brain and distinguish a reliable from an unreliable part in a generic way. It’s all one.
November 30, 2009 at 11:22 pm
And following up on Kay’s remarks (if I understand them properly), the point here isn’t about suggesting that we shouldn’t do phenomenology or the descriptive analysis of the structure of consciousness. The point is about opening up what is permissible within the field of philosophical investigation and inquiry. You might recall the posts I wrote a long while back on the hegemonic fallacy:
http://larvalsubjects.wordpress.com/2009/01/13/the-hegemonic-fallacy/
http://larvalsubjects.wordpress.com/2009/02/05/object-oriented-philosophy-what-is-it-good-for/
http://larvalsubjects.wordpress.com/2009/07/27/self-reflexivity-and-the-hegemonic-fallacy/
Given that these posts were written months ago and my views have developed quite a bit since then, I would have to go back through them to see if I still endorse the hegemonic fallacy in precisely the terms that it was presented there. Nonetheless, I think the spirit of the hegemonic fallacy still holds. When one term or type of actor comes to overdetermine all the others philosophy suffers. In this case, that term would be conscious awareness and direct access. Where conscious awareness or direct access to mind is treated as the foundation of philosophical thought and demonstration, then it becomes impossible to talk about things like physics, neurology, chemistry, biology, and etc. Why? Because these things are not directly given to consciousness. Insofar as this foundational move requires us to discuss everything in terms of how things are given to consciousness, discussions of something like neurology and biology are treated as dogmatic. Only description in terms of the givenness of consciousness is legitimate, e.g. only phenomenological analysis becomes admissible in philosophy.
There are a whole series of what I would call “hegemonies” in philosophy, each of which posits its own form of givenness or foundation (what we are warranted in talking about) and each of which excludes a whole host of other things as “dogmatic” (“baselessly speculative”).
* Phenomenology requires us to trace everything back to the givenness of consciousness, intentionality, or lived experience, and rejects any assertion not based on givenness to consciousness as dogmatic.
* The linguistic turn in both Anglo-American and Continental philosophy treats language as given or immediate and requires us to analyze all dimensions of experience in terms of how they are produced or structured by language and treats any assertion of language independent entities as dogmatic or baselessly speculative.
* Hermeneutics treats historically and textually informed consciousness as what is given or immediate, requiring us to analyze the manner in which our experience of the world is a product of how we inherit a particular tradition composed of texts and treats any discussion of the world apart from that textual history as dogmatic and baselessly speculative.
* Naturalism treats the physical material world as what is given or immanent, requiring us to trace all formations back to some sort of biological, neurological, or material explanation and treats all other entities (cultural phenomena, norms, texts, signs, etc) as mere epiphenomena.
The list could go on and on and certainly I’m not doing any of these positions the justice they deserve in this brief post. The point is that with each of these hegemonies a whole range of things is dogmatically excluded based on the philosophical decision to treat one of these elements as the privileged ground of everything else. The point is not to ban phenomenological description, semiotics, linguistics, economic analysis, naturalistic analysis, and so on In my experience online I’ve noticed that people have a tendency to either think loosely (not taking seriously what they themselves are claiming, or to think in binary terms. For example, if I criticize Lacanian thought for illegitimately excluding the neurological through its valorization of the symbolic or language, people immediately seem to think that I’m suggesting that Lacanian thought should be replaced by neurology. When I then turn around and start using Lacanian concepts and categories, people think there’s some sort of inconsistency here: “But wait, I thought you rejected psychoanalysis!” This has been, in particular, a problem in my dialogue with Asher who not only knows next to nothing about psychoanalysis but who has a rather poor understanding of neurology, in my view, as well. This is not to say I’m an expert in neurology, but Asher seems to have the idea that if something takes place in the brain it is innately or genetically caused rather than a product of how the brain has developed (i.e., he forgets that the brain is plastic and that much of its structure is the result of an interaction with the environment, part of that environment being other people and culture). But the point here is not one of choosing neurology over psychoanalysis. Rather, for example, it is a question of limiting the pretensions of psychoanalysis to make room for neurology and limiting the pretensions of neurology to make room for psychoanalysis, semiotics, and a whole host of other real things that make up the furniture of the world.
An important qualification here is that often these exclusions will not be stated explicitly by practioners of a philosophical school of thought or in the texts of these thinkers. You will seldom (though it does happen) read a phenomenologist explicitly excluding something like neurology or genetics (or alternatively, Foucaultian power and discursive systems, or Levi-Straussian social structures, etc). Rather, these exclusions generally have to be inferred by their absence in certain discourses and by the strong reaction that takes place among proponents of the philosophical school when findings from these excluded domains are evoked in argument. Moreover, these are generalizations about philosophical schools, dominant tendencies, not absolutes. In phenomenology, clearly Merleau-Ponty, for example, is better at taking into account social structures, power, language, biology, neurology, and so on than Husserl or Heidegger. There are gradations here of better and worse. Clearly Arian Bazaan in the world of Lacanian thought is better at taking into account neurology (she is herself a research neurologist in addition to being a Lacanian analyst) than J.A. Miller who never misses an opportunity to lampoon such research. There are all shades of gray here, but it’s important not to render discussion of these hegemonies and how they function invisible by virtue of evoking shades of gray in a reactive and knee-jerk fashion (a rhetorical ploy that often functions to deny the discussion of very real institutional realities).
In some of these discussions I think one has to have a pretty solid grounding in philosophy to understand them. Folks read a philosopher and think that philosopher is just pulling all sorts of wild claims out of thin air and are utterly baffled as to why anyone would claim such a strange thing. This arises from not understanding how philosophy is an ongoing conversation that is responding to its own history and the claims that have dominated that history. Thus you read me when I critique claims about direct access and think I’m saying something absurd like that we don’t have any access to our thoughts and wonder why anyone would say such an odd thing. What is invisible to you is that such an argument is responding to very specific epistemological claims about the foundational status of consciousness with respect to what is well grounded philosophically and what is not well grounded philosophically, and that if we wish to talk about things like technology, neurology, biology, physics, climate change, etc., philosophy requires a sort of medical therapy to disabuse itself of these assumptions, opening up inquiry into these other areas to which we do not have direct and unmediated access. However, if this is to be possible it is first necessary to show that our access to our own minds is no more immediate or well-founded than our access to other objects in the world, thereby dethroning consciousness and first person descriptivism from its privileged place. This line of argument, however, only makes sense in a field of philosophy dominated by phenomenology as its paradigmatic form of thought claiming to have “the most rigor and certainty”. Once this notion is overturned we can continue doing phenomenology with the caveat or asterisk that our phenomenological claims are fallible, that they are not a privileged foundation, that they are every bit as “speculative” as our claims about brains, biology, chairs, rocks, stars, and so on, and that these other modes of analysis are both legitimate, non-dogmatic, and can occasionally and often overturn certain claims in phenomenology.
November 30, 2009 at 11:59 pm
Thanks for your response, Levi. It might be useful to note that cognitive self-awareness has played nearly opposite roles in empirical psychology (my background) and in foundationalist epistemological philosophy. For a long time psychological researchers renounced all direct access to consciousness, relying entirely on the study of observable behavior. Only within the past 30 years or so have “thinking aloud” self-report methods gained acceptance and achieved statistical validation. There is, of course, acknowledgment that the technique works better in some aspects of cognition than others.
In my example about naming my childhood neighbors, I could talk through the way I arrive at the names: picture the house to the right, then the one to the right of that, then the next one, then skip down to the far end of the block where my pals Drake and Tim lived, then work back toward my house, etc. Talking through the process can provide useful information about how people retrieve memories, even if they can’t tell you precisely where the memories are stored in their brains and even if their memories aren’t perfectly accurate. Or I could ask you to calculate 41 times 18 in your head. If you just tell me the answer I learn very little about your thinking. But if you think aloud I can learn something about your cognitive strategy. E.g., you might say that 41 x 18 is (40 x 18) + (1 x 18). 40 x 18 is (40 x 20) – (40 x 2), etc. Think-aloud protocols like this are quite useful in evaluating thought processes and how people break down problems into simpler components.
As you point out though, people aren’t very accurate at giving their rationales for taking emotionally-charged actions. This sort of inaccuracy undermines the thesis that we have immediate access to all of our thoughts. There are plenty of other examples of reflexive failure. Distinguishing accurate from inaccurate self-cognitions, as well as exploring possible reasons why, is a major area of empirical research. As a result, looseness and fuzziness is systematically and incrementally reduced.
“When I argue that we do not have direct access to the contents of our own mind I am taking the central claims of other philosophical positions seriously and at their word and showing why they fail to hold up under scrutiny.”
Right. So it’s enough to demonstrate that people don’t have direct and accurate access to all of their thoughts. It’s not essential to claim that we have no direct access. We’re in agreement here.
If I’d been asked to think aloud prior to writing my post, here’s what I might have said: “Levi claims no direct access or relation of minds to themselves. Why is he making this (imho) obvious overstatement? I know that in his ontology Levi asserts that objects don’t have direct access/relation to each other. Maybe it’s part of this ontology to reject direct access/relation of objects to themselves as well. If I can subtly point out that minds do have at least some direct self-access, however limited, perhaps Levi will rethink his position on no self-access of objects, which I think should do no substantive damage to the rest of the theory.” It seems that I was drawing an incorrect inference about your ideas here, and so taking unnecessary corrective actions. Still, I think we’re in basic agreement on what minds can and cannot do in accessing their own contents and processes.
December 1, 2009 at 12:08 am
John,
That’s a good point about cog sci. I think this is why the generality of the hegemonic fallacy or the critique it tries to develop is important. On the one hand we have phenomenology dogmatically rejecting anything that can’t be traced back to consciousness, while on the other hand we have cog sci rejecting descriptive analysis altogether. The problem with the latter approach is that the cog sci theorist is going to have all sorts of assumptions about the nature of experience (how it is structured and lived) that are very likely unfounded. In AI, for example, years were wasted trying to develop artificial intelligences where memory and semantic structures were organized around genus/species trees and “semantic dictionaries”. Had they made the detour through phenomenological consciousness of lived experience they would have very quickly seen that consciousness is not organized in this way. This is why, I take it, a number of philosophers of mind and AI folks have recently become interested in naturalized forms of phenomenology. The point, then, is that the critique works in both directions.
Returning to your example of names, there’s still the question of why these names come to you when you seek to recall the names of your neighbors and why they come to you in this order. As you list those names it would also be interesting if you forget this or that particular name. The Freudian, of course, will claim that there is no randomness or accidents in cognitive life. Lacan echoes this point later in his post-script to the seminar on the Purloined Letter and in seminar 1 or 2 (can’t remember which). The idea here would be that the order in which you remember the names along with those names you forget has some sort of significance of which you’re not aware. I take it that this is more a heuristic thesis than something that could be demonstrated or shown. The idea would be to treat anything that leaps to our minds as possessing intentionality regardless of whether or not we’re aware of it.
December 1, 2009 at 12:55 am
Fascinating to think about gaps and errors in a way that links cog-psych and analytic models of memory and forgetting. During my afternoon run I was trying to do the old-neighborhood memory task. My overall strategy was fairly systematic, but as I stopped mentally at each house I found other associations coming to mind. So I thought about my next door neighbors the Littles, then their first names (Dewey and Dot) as being paradigmatically suburban. I remembered their dog, a boxer, but couldn’t remember his name at first. Only later, after remembering the name of the horse across the street (Rosebud, believe it or not), did the Littles’ dog’s name come back to me: Hector. Then I headed next door to the Heydeckers. I remembered that they were German, which got me to wondering whether their family name might have been Heidegger before being anglified at Ellis Island. Then I remembered that the oldest Heydecker son was very effeminate: he eventually moved to SF to become a hairdresser. For the next house, the Demas, I remembered shoveling their sidewalk from a heavy snow and cheapskate Mr. Dema paying me only a dime because I didn’t do his driveway, and how I thought at the time about shoveling the snow back onto his walk during the night. And so on up and down the street. To explore these childhood associations would be great for its own sake analytically, but it might well turn out that the associations either enhance or inhibit memories of names, depending on the affective charge of those associations.
Regarding AI, my somewhat disappointing experience in the field was that human-based empirical models of cognition weren’t as useful as one might think when building pragmatically useful devices like expert systems. This is because humans organize their knowledge in ways that compensate for their memory and processing limitations. Computers aren’t saddled with those limitations, so the programming can veer more toward machine efficiency rather than mimesis of human information processing.
December 1, 2009 at 1:04 am
From what I’ve seen these computational models encountered insurmountable problems in something as simple as removing a coke can from a desk (because of shifts in perspective when viewed from different angles) or in recognizing the difference between “my deer loves its salt lick” and “my dear adores PJ Harvey”. Phenomenological models of mind contribute a good deal to solving these sorts of problems and increasingly AI has shifted towards organic models of artificial intelligence to resolve these issues. It was a rationalist/Aristotlean model of thought that caused these problems in the first place. Had they taken the detour through phenomenological descriptive analysis this wouldn’t have been an issue, but because they began with rationalist assumptions about the nature of reasoning, truth, and thought they fell into all sorts of problems.
December 1, 2009 at 1:16 am
That’s a good point about cog sci. I think this is why the generality of the hegemonic fallacy or the critique it tries to develop is important. On the one hand we have phenomenology dogmatically rejecting anything that can’t be traced back to consciousness, while on the other hand we have cog sci rejecting descriptive analysis altogether. The problem with the latter approach is that the cog sci theorist is going to have all sorts of assumptions about the nature of experience (how it is structured and lived) that are very likely unfounded.
OK, I have to take exception to that. Any cognitive scientist, or at least any one that I know, would tell you that much of what goes on in developing hypotheses and ways to test them is a form of descriptive analysis, or introspection. We just require something beyond that as well, namely empirical investigation from a “third-person” (as Dennett calls it) perspective (usually, though not always, in the form of proper experiments). What’s more, phenomenological analysis has begun to make inroads into cog sci of late, though not always in particularly satisfying ways.
Anyway, I started to comment (before reading the comments) for a couple reasons. One, the study on the amygdala and moral judgments that Lehrer cites is probably a bad example. It has a bunch of methodological and interpretive problems. As an example of the latter, consider an alternative explanation: emotion is input into moral reasoning, instead of emotion doing the work and then reasoning doing the post hoc rationalizing. This explanation is further supported by the fact that people with amygdalar reasons are particularly bad at practical reasoning in general, not just moral reasoning. One could easily conclude that reason is blind without emotion, instead of that there’s emotion and then justification. When reading Lehrer, it’s best to remember that he’s not a scientist but a journalist, and that his evaluations of research are suspect (something he’s shown repeatedly in his books and on his blog).
More on topic, though, I think that there’s a reasonable sense in which we can say we have “direct access” to our mind, but to get to that sense, we have to take a step back and consider what it is, exactly, that’s a candidate for being directly accessed. I suspect that at some point in the distant, though perhaps not too distant, future, the conscious-unconscious distinction will undergo some major revisions, but for now, I think we can safely describe the mind as not one but two black boxes. The first black box, which is big and has multiple inputs (including itself and the other black box), represents the cognitive or adaptive unconscious (which is actually made up of other black boxes — we can probably distinguish three different types of unconscious processes: sensorimotor, automaticity, and the sorts of cognitive stuff we usually associate with consciousness). The second box, representing consciousness, is much smaller, and has only one input source, the big black box. Given this model, we can speak of two types of direct access: the first is direct access to the rest of our bodies and (if we include the retina, the ear, etc., in this black box) the outside world. We use this access to do all sorts of stuff, and we have direct access to the output of this stuff too. The big black box, then, which is as much “us” as the little one, has direct access to a lot of sutff. The smaller black box, on the other hand, has direct access only to the output of the bigger black box: direct access to heavily processed perceptual information, to the input for deliberation (which may be conscious to an extent, but which always had underlying unconscious processes operating within it), conscious volition, etc. We might further say that the little black box’s “direct” access is mediated by attention. That is, what it has direct access to at any given moment is contingent on attentional processes which are almost entirely unconscious.
It might be further reasonable to describe the access of the little black box, consciousness, as fragmentary. This is why introspection, and descriptive analysis, is so limited. We can’t reliably reverse engineer all the processing that went into the input that the little black box has access to, because we only get bits and pieces of even that heavily processed information. Descriptive analysis can overcome some of its limitations by describing similar (“the same” is a judgment that may be beyond it) experiences in a variety of contexts, from a variety of perspectives, as phenomenological analysis does, but it still remains fundamentally limited in that even an infinite number of contexts would fail to capture all of what the little black box doesn’t have access too.
December 1, 2009 at 1:19 am
Yes, you’re right: these AI models were often deep but narrow, good at manipulating expert knowledge but poor at the boundaries where common knowledge kicked in. Still, the systems were often pretty darned good at their assigned tasks. Limited deployment in the real world was more often a function of organizational than technical constraints. Still, plenty of simple AI gadgets populate the world behind the scenes. The Google inference engine is a pretty simple AI scheme, but people have adapted to it rapidly.
December 1, 2009 at 1:39 am
Chris,
You describe a number of the problems I have with Lehrer. In his writing Lehrer has a marked tendency to nativism, tending to claim that our moral reasoning is biologically innate. He compares moral reasoning to aesthetic judgments of taste. The first major problem with this line of argument is, as you point out, methodological: It fails to take into account the role of other people (which can’t be captured in fMRI situations). The second is that he fails to recognize the role of learning and development in our moral reasoning. He seems to think that judgments of taste are innate or “hard wired”, yet in comparing moral judgments to judgments of taste he opens the door to all sorts of important criticism. When I was a young lad there were all sorts of foods, drinks, and works of “art” and “literature” I hated and adored. Yet as I grew I came to appreciate things that previously I would have loathed and that I once loved I now detest and couldn’t even stomach. Clearly something takes place in these transitional periods at the level of the amygdala that cannot be treated as innate or genetic. I think this severely undercuts a number of Lehrer’s arguments.
That aside, I do think you’re somewhat missing the point of Lehrer’s inferences and arguments. His point is that the decision is already made and the action already executed by the time we reach the stage of providing the rational reasons for that judgment. Here I think Lehrer is right. The major point of dispute would be in the observation that the framework of that “immediate reaction” is not something biologically hardwired, but something that can be changed as a result of development and learning. Nonetheless, the conclusion that follows from Lehrer’s observations is that 1) emotion plays a key role in moral reasoning, despite what so many have argued in the tradition of moral thought where we’re supposed to exclude the affective dimension (and this has been well documented by cog sci folk who are actual practioners and not science journalists), and 2) that moral reasoning is not a deliberative process where reasons precede the decision. This entails that ethical theories and moral psychology that begins without taking account of these sorts of things will be hopelessly distorted and inaccurate.
Moreover, it will pose the wrong sorts of questions. Lehrer gives, for example, the striking example of studies where subjects were asked whether it was ethical for a train conductor to divert a train on to another track, killing one man, when it would save the lives of five men working on another track and whether it was ethical to push an obese person off a train platform to stop an oncoming train when it would save the life of five men uptrack. From a Kantian and a utilitarian perspective, both cases are structurally identical, yet the majority of people say the first case would be okay whereas the the second would be unethical. This is interesting as it suggests that our ethical responses are largely dependent on whether or not we have a personal identification with the person involved in the situation (something I think is aptly demonstrated by some of the ugliness that’s occurred on this blog and other blogs in the last couple of months). This suggests that the real ethical question is not that of how it is possible to ground ethical reasons that insure that folks will keep contracts, not kill, etc., but rather how it is possible to extend the domain of the ethical beyond the personal relation to the domain of the anonymous. That’s a good nugget if you ask me.
As for your observations about cog sci and the manner in which it takes descriptive analysis seriously, I think you need to go back and reread John’s comment carefully. John was making a comment about the training that he went through. This training occurred years ago. And certainly within the context of cog sci theory and practice when he was undergoing training, what he says is correct. A lot of things have changed since then, but this wasn’t always the case.
December 1, 2009 at 2:04 am
Somewhat famously,Wittgenstein says in “A doubt without an end is not even a doubt>(#625). This bears some relation to Descartes but it does not include the subject, the I of the cogito and this then perhaps escapes from the critique of Lacan. One could say in the context of this blog (maybe), doubting is a thing/event for which the notion of self knowledge or subjectivity or maybe even knowledge itself is superfluous. Rather what we have is something like a local manifestation of becoming and one which is distinctly impersonal even as it occurs “personally.” This possibility is important to me as a bridge between the epistemological and the ontic, the “flat” and the “human.” As I said, maybe.
December 1, 2009 at 2:13 am
I interpreted Chris’s point about descriptive analysis as having to do not specifically with cog psych but with scientific hypothesizing in general. And I agree: even during the heyday of behaviorism (even before my time, incredibly), scientific hypotheses didn’t just fall out of the data analyses like the hard-core positivists of the time might have asserted. Instead, scientists consciously reflected on the empirical evidence. And then these hypothesizers had to be self-aware about those reflections in order to write them down. So there was an internal contradiction within psychology that eventually got overthrown by the so-called cognitive revolution in psychology. Or maybe I’m just too damned old to follow Chris’s train of thought properly. In those long-ago days of my formal schooling, phenomenology was pretty much limited to JJ Gibson’s ecological perception and Roger Barker’s work on behavior settings.
Here’s another example of affect interfering with accurate self-reflection. In a social psych study of romantic relationships, one group of subjects were asked to introspect about specific reasons why their relationships were going the way they were, then to rate their satisfaction with the relationship. The other group was asked to rate relationship satisfaction on gut feel, without the introspection about factors that might contribute to their ratings. Subjects were followed for some number of months, to see which couples were still together and which had broken up. The satisfaction ratings of the “gut feel” subjects were predictive of relationship outcome. The “navel gazers,” on the other hand, showed no correlation between satisfaction and relationship persistence. It seems that the introspection interfered with the affect, resulting in conflicted self-awareness and inaccurate ratings. Empirical psychology marches on!
December 1, 2009 at 3:02 am
John’s interpetation of what I meant in my comments about descriptive analysis are correct. And I don’t think the position of cognitive scientists on this matter is dogmatic. We have good reasons to demand more than description, as the content of this post shows.
On the moral stuff. I’ve been doing moral psych research for about 4 years now, and long ago realized that, from a methodological standpoint, the trolley and footbridge problems are deeply problematic, to the point that conclusions drawn from their use are always either unwarranted or flat out wrong. I am not the only one who feels this way, and as a result, the field is slowly moving away from reliance on comparing dilemmas. And just to show how those two dilemmas are not structurally identical, even from a Kantian perspective, consider the possibility of self-sacrifice, which occurs only in one. What’s more, the dilemmas require a great deal of suspension of disbelief (participants often question whether throwing a guy on the tracks will work, e.g.).
What I really meant by the methodical and interpetive problems with Greene’s work, though, has to do with what they say is going on. Greene’s theory, which Lehrer relies on, is that there is an affective and a cognitive component to moral decisions (with only the cognitive component being active in the trolley case). On the one hand, this shows that in “impersonal” moral decisions (Greene’s distinction), the “before we’re aware” argument doesn’t work. On the other hand, there are plenty of equally plausible alternative explanations (even if we forget that the methodology makes explaining the results nearly impossible) that don’t require us to accept that the decision is made unconsciously (via processes in the amygdala, e.g.) before we’re aware of it. Certainly we react before being aware, and this is important, but there may be many cases when the reaction is merely input into a more deliberate moral judgment. This is perfectly consistent with the data, at least.
December 1, 2009 at 3:08 am
Chris,
I look forward to learning more from you about why empirical psychology is completely irrelevant, how philosophy already mapped these things, and why we can rely completely on the self-evidences of our introspection. It is a tremendous relief to discover that our phenomenological analyses of our decision making processes are incorrigible and we have nothing to learn from fMRI and these sorts of studies. Thank god I don’t have to leave my office!
December 1, 2009 at 3:25 am
Levi, I’m not sure which part of my comment you’re replying to. I don’t think anything I have said implies that either empirical psychology or phenomenological analysis is useless. As an empirical psychologist who was originally trained in phenomenological analysis (before leaving phil of psych), it would be weird if I did. I apologize if I appeared to do so.
In actuality, I think they both work well together, and independently are limited or ineffective (I thin phenomenology without empirical psychology is more limited than empirical psychology without phenomenology, though).
My beef with Greene’s studies and Lehrer’s use of them is mostly limited to those studies. I think they’re next to useless for moral psychology, and that much better empirical work exists. And much of that work is consistent with conscious deliberation playing a role, particularly in difficult or unusual moral judgments (many common judgments are likely automatic).
I will admit a deep skepticism of cognitive neuroscience, and especially of social neuroscience. We just don’t know enough about the brain to draw firm conclusions in most studies, and our technological and methodological limitations are immense.
December 1, 2009 at 3:34 am
One more thing: I agree with you about most of the points in this post, as I think my first comment shows. I have long argued, on my blog, on paper, and in the classroom, that we are infact “strangers to ourselves,”that consciousness plays a tiny though important role in our mental lives, and that the focus on consciousness in philosophy (historically, as well as in both the contemporary Continental and Anglophone/Analytic traditions) is unfortunate. Over at my old Science Blogs site, there is a link (under the S) to a picture of a monkey riding a tiger. That comes from one of my favorite quotes on consciousness (easy to find on the web). I think direct access is a reasonable description of some aspects of our mental lives, including our unconscious lives (limiting the word direct and access to consciousness is part of the problem), but I don’t think we could, even given infinite time, use our conscious experience to adequately characterize the bulk of our mental lives. And that is why I have a job.