Introduction: a reminder of the PreSocratics...the pattern of reducing qualitative differences to quantitative differences.
A definition of reduction in science, examples
Two models of intertheoretic reduction: Nagel's Model (and "glitch"); Kemeny and Oppenheim's Model (and glitch)
Unity of Science Program - and glitch
Natural Kinds and their use in strengthening the position of Antireductionism in the context of psychology
The Impact of Antireductionism
Argument against weak supervenience
Impacts and implications: supervenience (apparently) fails to sustain a strong unity of science program
Reductionism in the PreSocratics
Thales-all things are water
Anaximander-all things stem from the boundless
Anaximenes-all things are air
Further, he showed how, unlike Thales and Anaximander, causal mechanism by which one substance with qualities xyz changes into another substance with qualities abc (though condensation & refraction)
So, all qualitative differences are reduced to quantitative differences.
We want science to tell us what is really going on in the world-what's in it, how it got that way, and where its going…(emphasis added, CE)
Some philosophers of science claim that how we measure progress in science comes down to the degree to which the several sciences are unified into one interconnected network of theories (that it's a single thing whose parts fit together into a coherent pattern)…
Instead of there being more basic objects and properties in the universe, there are fewer. And that is good, because the fewer the basic objects and properties, the lower our risk of error in doing science (there is less for us to be wrong about). The last sentence leads naturally to a principle of scientific method often called Ockham's Razor.
Reductions in History:
(1) Thermodynamics in physics is reducible to statistical mechanics.
(2) In biology Mendelian genetics is reducible, or so some have claimed, to molecular genetics.
(3) The reduction of chemistry to physics.
Partial reductions:
How is reduction carried out?
Two models of intertheoretic reduction:
Nagel's model of reduction was tied to the positivist conception of scientific theories in the sense that reduction was applicable only to theories that were formalizable as first-order languages in the way the positivist conception demanded (the theory must have a set of core theoretical laws, and it must have a vocabulary distinguishable into two mutually exclusive classes of terms: theoretical and observational).
*assuming that scientific theories can be so formalized, Nagel's model of reduction is remarkably clean in principle but extremely sloppy to pull off in practice…
Let T1 and T2 be formalized theories. Then T2 reduces to T1 if and only if the following conditions hold:
1. The Condition of Connectability: For every theoretical term 'M' that occurs in T2 but not in T1, there is a theoretical term 'N' that is constructible in T1 but not in T2 such that:
for all objects x, x has M, if (and possibly only if) x has N (bridge law)
The bridge law serves as a bridge to link two distinct linguistic expressions in two distinct theories.
Example: Researchers believe that the gene associated with the disease cystic fibrosis resides on chromosome 7. Suppose further that eventually the branch of specialized medicine that treats diseases like cystic fibrosis is reduced to some form of supergenetics. …for all organisms o, o has the cystic fibrosis gene if and only if o has DNA nucleotide sequence Z-2134 at location X on chromosome 7…
So, to recap Nagel's method of reduction: (1) First we must logically connect the terms of the respective theories by constructing bridge laws; (2) second we must be able to drive logically the laws of the reduced theory from the laws of the reducing theory plus the bridge laws.
Glitch in Nagel's Model:
The reduced theory is often false in the sense that it only holds good in special circumstances.
Example: read from book pg. 87
Argued that the Condition of Connectability in Nagel's model of reduction is almost never fulfilled in what are otherwise considered successful reductions (the reductionist is almost never able to construct term-by-term bridge laws, for the holistic interdependence of theoretical terms defeats the attempt to do so…the bridge laws become so huge and unwieldy as to be useless, if not meaningless)
Kemeny and Oppenheim argued that what matters in a reduction is that the reducing theory be able to explain any observational data that the reduced theory explains, that the reducing theory be in some genuine sense different from the reduced theory, and that it be at least as systematic as the reduced theory. Their model also did not try to provide necessary conditions for a successful reduction, but only sufficient conditions.
Let T1 and T2 be formalized theories. T2 is reducible to T1 if:
Condition (1) ensures that there is an ontological point to reduction. Condition (2) makes explanatory power a key factor in the evaluation of intertheoretic reductions. Condition (3) is required, said Kemeny and Oppenheim, to rule out wrong-way reductions. For example, reducing psychology to astrology is forbidden because astrology is theoretically more disorganized, more unsystematic, than psychology.
Glitch in Kemeny and Oppenheim's Model:
The flaw in the model is the third condition.
Some reductions involve reducing a simpler theory to one that is more complicated and less well systematized precisely because the reducing theory doesn't oversimplify the target phenomena where the reduced theory does oversimplify the target phenomena.
Example: Chemical theory to Quantum theory…pg. 88
*In summary, what really motivate the reductively minded practitioner are causal and ontological issues, not issues of explanation and systematization. Instead of having reductively minded practitioners care about what causes what out in the real world, Kemeny and Oppenheim have them care mainly about how tidy and spiffy their theories are. Like Thales, Anaximander, and Anaximenes, most scientists are ontologists at heart. They wish to understand the cosmos, and to understand it as a cosmos, and that means that they wish to understand how its parts are unified into a cosmos.*
Developed by philosopher, Hilary Putnam, and once again, Paul Oppenheim
Core idea: There are no special sciences, where by special science means a science whose fundamental concepts are somehow incapable of being connected with or related to the fundamental concepts of at least one other science.
If the universe began from a singularity that exploded, as the big bang claims it did, then the universe is bound to be a completely homogeneous entity; that is, there would be no ontological grounds for some parts of the universe being so qualitatively distinct that they can't be fit in theoretically with any other parts of the universe (everything must be singular).
Special science is a science that destroys the homogeneity of the universe in virtue of their specialness. The laws that would presumably be at the core of any special science that investigated such phenomena are called nomological danglers.
Nomological danglers are rejected for three reasons:
Argument for the Unity of Science:
Putnam and Oppenheim presented an argument for the Unity of Science that involved mereological premises: principle of evolution & principle of ontogenesis
Principle of Evolution-the universe evolved from smaller to larger in terms of organization; so, for any level of organization in nature, there was a time in the evolution of the universe when there were systems of that level but no systems of any higher levels of organization.
Principle of Ontogenesis-there was a time in the evolution of the universe when the system did not exist but some of its parts on the next lowest level of organization did.
Glitches in Putnam and Oppenheim's premises and consequently the Unity of Science:
The principle of evolution assumes that there is a natural and unique way to decompose the universe into levels of decreasing organization, that there is one correct way to do this. However, there are different ways of cutting up a system into parts. It all depends on what your purposes are in decomposing the system.
The principle of ontogenesis assumes something even more questionable: that it makes sense to speak of a system's parts existing before the system itself exits. But there are notorious identity problems with this way of proceeding. In other words, the prior existence of the system is a necessary condition for the existence of any of the system's parts.
Putnam came to reject the Unity of Science Program only a few years after he'd written with Oppenheim. Philosopher, Jerry Fodor, advanced Putnam's rejection of the Unity of Science.
Putnam made his rejection of the Unity of Science turn on the alleged nonreducibility of concepts within theories. This is what we mean by antireductionism.
If antireductionism is right then psychology, for example, is a special science in the strongest sense possible. In other words, certain concepts within psychology are nonreducible.
Fodor argued that there is a theoretical discontinuity between the sciences that supposedly meet in the reductive hierarchy at that level of organization. The disruption in the fabric of nature is where psychological phenomena meet neurophysiological phenomena.
Natural kind-A set of entities capable of having an essence forms a kind; and , if the entities in question are naturally occurring entities, the kind is a natural kind.
What natural kinds allow a practitioner to do, among other things, is to cut the domain of inquiry along its natural seams.
Like any other domain of inquiry, psychology seeks to uncover the genuine casual relations between the genuine natural kinds that form the seams of the psychological level of phenomena within the universe.
Psychological laws of nature will relate natural kinds to each other, natural-kind properties to each other, and natural-kind events to each other. But that means, argued Fodor, that psychology cannot be reduced to any theory or collection of theories from any sciences below it in the hierarchy presumed by the Unity of Science.
Fodor demonstrates how psychology cannot be reduced Nagel's Model of reduction (does so by way of the multiple realization argument):
When we understand how it is that psychological states and properties of organisms are physiologically realized in the nervous systems of those organisms, we shall see that the terms constructed in the reducing theory to form the bridge laws (Nagel's Model), in many cases, will not denote natural kinds.
The bridge laws, in many cases, will not denote natural kinds because a given psychological state is multiply realizable in the physiology of a typical nervous system.
Example: The psychological state of thinking of the first time human beings landed on the moon. Suppose two people are thinking of this. This is a mental state functionally characterized. That means that there doesn't have to be anything, not one single thing, similar in your brain and my brain in virtue of which we are both in this same mental state. That is why Fodor's argument against reductionism gets off the ground: The multiple physical realizability of mental states will result in bridge laws that aren't laws at all because their physiological sides do not pick out natural kind properties, objects, or events in the physiological domain of nature.
The Impact of Antireductionism:
*The Putnam/Fodor multiple realization argument created much consternation in philosophical circles. Some reductionists tried to duck its impact with a "so what?" response. Other philosophers jeered at the argument's use of the concept of natural kinds on the grounds that there are no such things; or worse, on the grounds that there are natural kinds but we get to say what they are, not nature (and so, the last half of such a bridge law is as natural as you want it to be). Still other philosophers took the argument to heart and said good-bye to reductionism. There is at least one special science after all-psychology.*
Supervenience was originally designed to save that which is noncausal, nonreductive physicalism from the aftereffects of the multiple realization argument. The model of supervenience discussed here is of philosopher, Jaegwon Kim.
The original purpose behind supervenience was a metaphysically deliberate one: to construct a concept of asymmetric dependence under which the determinative relation represented by the concept in question is:
Physicalism denotes what used to be called materialism, the view that the universe is ultimately an entirely physical system. What physicalism implies is that a mild version of the Unity of Science holds good: Ultimately there are no phenomena in the universe which cannot be understood in terms of the concepts of physics.
So, in a way, supervenience was called on to salvage a toned down version of the Unity of Science, a version in which the respective sciences line up in order of determinative dependency without lining up in reductive order
Supervenience is a relation that holds between natural groupings or what Kim called families of properties. Example: One way of understanding this is to note that competent supervenience theorists rarely talk about a particular property supervening on another particular property…they always talk in more general terms such as, "the mental supervenes on the neurophysiological or allergic phenomena supervene on immunological phenomena". Or, mental properties and events (as a family of such) supervene on neurophysiological properties and events (as a family of such).
Let the family of psychological properties attributable to higher organisms be represented by M* (the point of the * after the M is to indicate that we include n M* properties formable by applying simple logical operations to the basic properties of M*, operations such as conjunction, disjunction, and negation. Let the family of neurophysiological properties attributable to higher organisms be represented by N*. Then the psychological supervenes on the neurophysiological if and only if the following holds:
necessarily, for any two objects x and y, if x and y have exactly the same N* properties, then they have exactly the same M* properties; and, if x and y differ in M* properties, then they differ in N* properties.
*Supervenience comes in different strengths: Global, Weak, & Strong. The latter example was weak supervenience or moderate supervenience.
Argument against weak Supervenience:
Arguments can be heard that weak supervenience is not robust enough to qualify as an interesting resolution of either the traditional mind-body problem, or the Putnam/Fodor multiple realization argument.
The use of strong supervenience introduces a highly complicated, technical, and possible-world semantics
Supervenience as applied to Psychology or Antireductionism:
There is no difference in psychological properties without a difference in neurophysiological properties, but there can be a difference in neurophysiological properties without there being a difference in psychological properties.
In other words, the psychological can supervene on the neurophysiological in a one-to-one fashion: The same psychological state can be neurophysiologically realized in many distinct ways. The neurophysiological varies if the psychological varies, but the psychological does not necessarily vary if the neurophysiological varies.
The determinative dependence is asymmetric.
Is Asymmetric Covariance a relation strong enough by itself to sustain a watered-down Unity of Science program?
Kim states, "The results of work by many supervenience theorists in recent years indicate that asymmetric covariance is logically independent of asymmetric determinative dependence: Two families of properties can stand in a relation of supervenient asymmetric covariance, and it can be the case that neither is determinatively dependent on the other."
Determinative dependence would surely be one minimum requirement of any Unity of Science program
If the various sciences cannot be lined up in reductive order, then at least they should line up in order of determinative dependence-they should, that is, if the Unity of Science is to have even the slightest interest.
If Kim is correct, then it is a separate task to show that a form of determinative dependence underlies a given supervenience relation. Simply pointing out that supervenience holds will not be enough.
Showing that a Form of Determinative Dependence Underlies a given Supervenience Relation:
This model applies to different domains of inquiry--
If A supervenes on B and B supervenes on C, then A supervenes on C. The presence of an intervening third family of properties in the supervenience chain from A to C may serve to illuminate the determinative ground of the supervenience relation of A on C.
Example: starting point-(1) allergic properties supervene on immunological properties (a supposition with much evidence behind it).
*Determination at that point is pretty much mereological in nature
Example: In cognitive psychology and neurophysiology, it is presumably an open question subject to controversy just what kind of determinative relation grounds the supervenience of the psychological on the neurophysiological. The behaviorist, for example, might argue that the determinative relation in question is semantic while the funtionalist pyschologist would, of course, argue otherwise.
Impact and Implications of Supervenience:
What the demand that we distinguish supervenience from its ground amounts to is an admission that supervenience does not itself constitute an explanatory basis for anything.
The existence of a supervenience relation between property family A and property family B suggests that there is a determinative priority between the two families of properties; but it does not indicate the specific nature of that determinative relationship (is it mereological, causal, semantic, or some other kind of determination?).
Because it does not indicate the specific nature of the determinative relationship, supervenience does not by itself count as an explanatory relation between the two families of properties. It only suggests that there is some determinative relation between the property families in question that is sufficient to underwrite an explanatory relation between them; but the kind of determination may be different in different domains as mentioned in the preceding paragraph.
*Klee concludes stating, "Nowadays, there is less acceptance on the part of philosophers of science of such a monolithic conception of scientific explanation."*
Ess's comments:
While supervenience is originally offered in the effort to save some (originally positivist) program of reductionism and thus a strong Unity of Science program -
in the face of the Putnam/Fodor "multiple realization argument" (i.e., the observation that the same mental state is apparently realizable or instantiated in different ways physiologically) -
at least the form of "supervenience" offered here - already characterized as weak supervenience - amounts to an admission that no such reductionism seems possible.
This admission comes in at least two places:
Beginning with the characterization of "asymmetric covariance" between psychological properties and neurophsyiological properties as the relationship of supervenience
- it is asymmetric in part because it is not (a) causal
and (b) reductive.