They Myth of Psychological Laws

This entry was posted on 3/26/2008 8:20 AM and is filed under Philosophy of Science,Psychology.

Trackbacks

• No trackbacks exist for this entry.

Comments

• 3/26/2008 8:46 AM Ken wrote:
  Check out John Anderson's textbook on learning and memory.
  Reply to this
• 3/26/2008 9:28 AM Anibal wrote:
  I like this passage of Stanislas Dehaene that has appeared as an answer to the Edge Annual Question 2008, in relation to the possibilities of psychological laws:
  
  "Although a large extent of my work is dedicated to modelling the brain, I always thought that this enterprise would remain rather limited in scope. Unlike physics, neuroscience would never create a single, major, simple yet encompassing theory of how the brain works. There would be never be a single "Schrödinger's equation for the brain" -and Dehaene continues- "Well, I wouldn't claim that anyone has achieved that yet… but I have changed my mind about the very possibility that such a law might exist. Very recently, however, Karl Friston, from UCL in London, has presented two extraordinarily ambitious and demanding papers in which he presents "a theory of cortical responses". Friston's theory rests on a single, amazingly compact premise: the brain optimizes a free energy function. This function measures how closely the brain's internal representation of the world approximates the true state of the real world. From this simple postulate, Friston spins off an enormous variety of predictions: the multiple layers of cortex, the hierarchical organization of cortical areas, their reciprocal connection with distinct feedforward and feedback properties, the existence of adaptation and repetition suppression… even the type of learning rule — Hebb's rule, or the more sophisticated spike-timing dependent plasticity — can be deduced, no longer postulated, from this single overarching law"
  
  In sum, perhaps there is such law in psychology and we have not to explain psychological phenomena in functional terms but in nomological ones.
  Reply to this
• 3/26/2008 12:44 PM Greg wrote:
  Just out of curiousity, what would be so weird or bad about these people using the word 'law' to include things like mechanisms? I mean, I understand there are differences between mechanical explanations in e.g. molecular biology and law-driven explanations in e.g. (certain fundamental) physics, but do those differences make a difference for the arguments given by the people you have in mind?
  
  (As you can tell, this is mostly asked out of ignorance.)
  Reply to this
• 3/26/2008 8:09 PM Pete Mandik wrote:
  Hi Gualtiero,
  
  I think that Chris Gauker gives some pretty good accounts of the origin of the myth of psychological laws in section 2 of the following paper:
  
  The Belief-Desire Law. (pdf document. 124K. Facta Philosophica 7, 2005: 121-144.)
  
  http://asweb.artsci.uc.edu/philosophy/gauker/BeliefDesire.pdf
  Reply to this
• 3/27/2008 8:52 AM gualtiero wrote:
  Guys, thanks for these great comments.
  
  Anibal, what you mention is extremely interesting but I'm sure it's not what your typical philosopher of mind who talks about laws was hoping for . They were hoping for a law about beliefs, desires, and the like.
  
  Pete, thanks so much for that reference. I just read Gauker's paper and I think it's great.
  
  Greg, your question is quite complicated. First, I think you already have part of the answer when you point out that laws and mechanisms point at two different conceptions of explanation. Why talk about psychological laws, if that's not how psychology explains (and philosophers of psychology have been pointing this out for decades)? Second, I definitely think that in at least some cases, whether you run certain arguments in terms of laws or in terms of mechanisms makes a difference to how the argument comes out. This is in part because of the kind of laws that philosophers of mind tend to have in mind, i.e., belief-desire laws. If you want to pursue this further, I would encourage you to read the paper by Gauker and Pete kindly linked to.
  Reply to this
  1. 3/27/2008 10:59 AM Anibal wrote:
     Ok, Gualtiero, i get the point.
     
     But, desires, beliefs and the like are not free-floating rationales they have to be "anchored", and in this sense, is ultimately relevant Friston´s insights.
     Reply to this
• 3/27/2008 9:58 AM Eric Thomson wrote:
  A couple of problems with this view.
  
  1. Weber-Fechner is a law. I can use such psychophysical laws as explananda. E.g., to explain why so-and-so didn't detect that change in light intensity. Sure, that law also can be explained, but so do many laws of physics (i.e., any that are not fundamental laws).
  
  2. Perhaps these misguided young philosophers are speaking a bit more colloquially, referring roughly to general theories with explanatory bite as 'laws.' For instance, they would consider Chomsky's theory a set of laws (indeed it seems to fit the bill quite well: provide some initial conditions and parameters and voila a DN explanation).
  Reply to this
  1. 3/27/2008 10:00 AM Eric Thomson wrote:
     Oops, I said such laws can be used as explananda, but I meant explanans.
     
     Reply to this
• 3/28/2008 8:28 AM Tony wrote:
  The idea of law is not exactly a clear one (van Fraassen 1989 and then all the recent work on laws in metaphysics, like Mumford 2004). One path here is first to divide metaphysical from non-metaphysical accounts of laws. The first are about what makes what happen. The second are about what we learn and write down. It seems to me to be extremely plausible that there are no psychological laws of the second kind. About the first, the enormous literature on the mental causation problem might either persuade you, or make you assume, that there are no laws that display mental properties making things happen. I think that's a mistake: I argue in my book (fortcoming with Columbia University Press, title "Mental Causation") that there's a way to see mental properties hooked up with various other properties to get laws of nature. The idea is kind of like Hempel's: instances of psychological properties don't reliably determine effects. So you have to qualify the laws to compensate for the differences. (This is where the difference between metaphysical and non-metaphysical conceptions becomes crucial: the "qualified" laws aren't anything that we could discover or write down.)
  Reply to this
• 4/5/2008 1:36 PM Cory Wright wrote:
  For what it's worth, there's some nice documentation of the exceedingly rare appeal to laws---much less appeals to laws qua explanans---in the psychological sciences. Teigen (2002, citation below) has a nice paper discussing a bibliometric study of abstracts from the PsycLit database during the last century (1900–1999). Having combed through more than 1.4 million abstracts, he found 3,093 citations of law---an average of 22 citations per 10,000 entries. (Unsurprisingly, the average dwindles from .0022 to .001 in the so-called Decade of the Brain, 1990-1999.) Teigen also found that the laws psychologists are most concerned or familiar with tend to be older, with the most commonly cited being Weber’s law (1834), Fechner’s law (1860), Donder’s law (1863), Emmert’s law (1881), the Yerkes-Dodson law (1908), Thorndike’s law of effect (1911), the Gestalt laws of perception (1923), Hick’s law (1952), Fitt’s law (1954), and Steven’s power law (1957). These latter results were collected anecdotally, but are corroborated by similar bibliometric studies. Roeckelein (1996, citation below), for instance, sampled 146 standardly used psychology textbooks published between 1885–1996 and also found a clear stagnation pattern in law citations. Carl Craver recently put me on to another such paper, by Roediger at WashU, which walks readers through the reasons as to why laws of memory and learning 'vanished'.
  But what about Gualtiero's question? Why is it that some philosophers talk as if psychologists explain in terms of laws? Or do they somehow use 'laws' to mean psychological mechanisms? One answer---perhaps the best but least satisfying---is that philosophers talk as if psychologists explain in terms of laws because they've merely learned a way of talking, and have failed to learn a new idiom. As for the second part, normally, to mean psychological mechanism we use the term 'psychological mechanism'; to mean law, we use 'law'. And so to mean psychological mechanism by 'law' seems egregiously idiosyncratic.
  ___________________
  
  Roeckelein, J. E.(1996). Citations of laws and theories in textbooks across 112 years of psychology. Psychological Reports 77: 163–174.
  Roediger, H. L. (2008). Relativity of remembering: Why the laws of memory vanished. Annual Review of Psychology, 59: 225-254.
  Teigen, K. H. (2002). One hundred years of laws in psychology. American Journal of Psychology, 115: 103–118.
  Reply to this
  1. 4/5/2008 2:49 PM gualtiero piccinini wrote:
     Cory, thanks for this extremely informative and sensible comment.
     Reply to this
  2. 4/5/2008 8:03 PM Eric Thomson wrote:
     Hi Cory--interesting references.
     
     There are a couple of issues.  One, do psychologists call their theories laws? Two <i>are</i> their theories laws?
     
     It seems the second question is more interesting, the first a sort of anthropology (and if you really wanted to go down that road, I'd bet there are even fewer references to 'mechanisms' in the same literature--without that comparison we can't judge whether the mechanism view or the law view is better in terms of frequency statistics).
     
     In practice, people in psych quite frequently talk about evidence, theories with varying degrees of support, and explanations. If they did a search for 'theory' in those texts, they would find a high percentage compared to physics. Perhaps the main difference is that basically no theories in psychology are well-enough established to merit the 'law' moniker. So then the criticism of 'law' talk should be seen as a special case of a general criticism of 'theory' talk, but theory talk saturates psychology. I might argue that (wannabe) laws are all over the psychology textbooks.
     
     I like to use the word 'law' sometimes just because Davidson's claim to the effect that there are no psychophysical laws sticks in my craw.
     
     Reply to this
• 4/6/2008 7:35 AM gualtiero wrote:
  Eric,
  
  I agree that you can use laws to explain. But unlike in physics, where laws are explained by more general laws, in psychology laws are explained by underlying mechanisms. The more general and fundamental form of explanation in psychology seems to be mechanistic, as a long strain of philosophers of psychology has recognized. (Sometimes philosophers of psychology say explanation in psychology is "functional", but the relationship between functional and mechanistic explanation is a story for another time. For now, let's just lump them together.)
  
  I don't think theories always lead to (putative) laws. In psychology, theories generally describe mechanisms, not laws.
  
  Here is my tentative answer to my question: Why do philosophers of psychology keep talking about psychological laws? Young philosophers of psychology talk about psychological laws because they take their cue from old philosophers of psychology; old philosophers of psychology talk about psychological laws because they are under the influence of old-fashioned (and largely superseded) views in the philosophy of science, such as the view that causation, scientific explanation, etc. require laws. (Davidson is definitely a good example of influential old philosopher who held some of those views.) Meanwhile, philosophy of psychology remains disconnected from their putative subject matter, i.e., psychology. So this talk of psychological laws is a mistake, which needs to be corrected.
  Reply to this
  1. 4/6/2008 10:32 AM Eric Thomson wrote:
     True, not all theories are laws, some are mechanistic models, as we often find in biology, especially molecular biology.
     
     Is there a good example of a successful mechanistic explanation in psychology?
     
     I agree that mechanistic explanations, as promoted by Bechtel, are used all the time in biology, and there are many successful examples of such explanations there. But some theories in biology have the form of laws even if they aren't called laws. Hodgkin-Huxely (HH) equations would be called laws if they had a more universal applicability. So biology includes both types of models. You can say that the HH model was subsequently explained in terms of mechanisms, but that is a different issue. The claim was there are no laws, and this seems wrong.
     
     Ultimately one reason biology has so few beautiul HH-like models isn't because we don't want them (we do), but because the systems are so complex, nonlinear, with so many parts interacting, it is only recently that we have been able to do the math with the help of computers (most of them do not have analytical solutions so must be simulated). So perhaps Bechtel's excellent historical analysis of mechanistic explanations in biology has that bias built in precisely because biology is just getting more mathematical with the advent of greater computer power.
     
     Where we can mathematize something, we will. The mathematization simply usually comes later in the evolution of a "mechanistic" explanation. HH, for instance, came many decades after people were reporting the voltages across cell membranes. If you look at biochemists, they love it when first-order enzyme kinetics apply to their system, as then they can apply simple "laws" to predict how things will behave. The Nernst Equation in biophysics is beloved by all neurophysiologists, is the basis for some elements of the HH equations, and it is derivable from physical laws.  Even the Kreb's cycle, a mechanistic account of ATP production, can also be explained in terms of the laws of enzyme kinetics.
     
     I'm not saying mechanistic explanations are eliminable, but that the law-like models and mechanistic models ultimately evolve in parallel.
     
     Reply to this
  2. 4/10/2008 7:06 AM Tony wrote:
     Gualtiero wrote, "(Davidson is definitely a good example of influential old philosopher who held some of those views.)" But it's bizarre to explain young philosophers' disposition to talk about psychological laws, since Davidson was so keen to prove that there aren't -- indeed, to show that explanation in psychology doesn't rely on laws!
     Reply to this
     1. 4/10/2008 8:34 AM Eric Thomson wrote:
        Yeah, as I said Davidson is one reason I like to say there are psychological laws (well, in addition to the fact that there are).
        
        Reply to this

Leave a comment