Abstract
One way that scientifically recognized properties are multiply realized is by “compensatory differences” among realizing properties. If a property G is jointly realized by two properties F1 and F2, then G can be multiply realized by having changes in the property F1 offset changes in the property F2. In some cases, there are scientific laws that articulate how distinct combinations of physical quantities can determine one and the same value of some other physical quantity. One moral to draw is that in such cases we have the multiple realization of a single determinate, “fine grained” property instance that is exactly similar to another instance. As simple as this moral is, it has ramifications for a number of recent discussions of multiple realization in science. Taken collectively, these ramifications indicate that multiple realization by compensatory adjustments merits greater attention in the philosophy of science literature than it has hitherto received.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
In principle, one might maintain that the relevant individual bearing trichromacy is the human visual system or the human eye. In any case, this does not affect the present issue of MRCD.
For an account of the levels invoked in clause (iv), see (Gillett unpublished) and (Wimsatt 2007).
This is a feature that the Dimensioned view shares with mechanistic explanation as described in Craver 2007.
Note that the language of “compensatory differences” should not be taken to imply that one instance of G is multiply realized by changing the realizers of G from the set F1-Fn to the set F*1-F*m. There appear to be cases in which this logical possibility is actualized, but this is not what is required by the language of MRCD.
One might worry that only Gillett’s Dimensioned view of realization allows for MRCD. The breadth of the foregoing examples is meant to challenge this. First of all, there are instances of MRCD countenanced by the “Flat” view of realization that Gillett rejects (Cf., Gillett 2002, 2003.) Consider ball bearings and roller bearings. Both kinds of bearing support radial loads. Call an individual ball or roller in a bearing a “rolling element”. Each rolling element in a bearing supports a specific radial load. The load a particular rolling element will support is a function of both its shape and the hardness of the material of which the rolling element is made. For many values of radial load, there are distinct combinations of shape and hardness that will bear that load. This is MRCD in a flat view, since the properties that realize the radial load capacity of an element are properties of that very rolling element, rather than properties of parts of the rolling element. The hardness of a rolling element is a property of the element and the shape is a property of the element.
Second, Craver 2007, like Gillett, offers a theory of compositional determination relations, only describing it in terms of mechanistic explanation, rather than realization. Thus, it is possible to adapt the Craver 2007 view of mechanistic explanation to provide an analysis of many of these cases. So, thinking of MRCD in terms of Craver’s mechanistic explanation, one might say that distinct combinations of entities and activities can lead to the same phenomenon by way of compensatory differences among the entities and their activities. So, to use an example to which we shall return, let a phenomenon be the propagation of an action potential at a given velocity. The entities of an axon, including its membrane and axoplasm, might engage in somewhat different activities of resisting and “capacitating” in such a way that distinct combinations of them can lead to the same action potential velocity.
Third, suppose one did have a theory of realization and multiple realization that did not cover the range of cases described above. In that event, one might reasonably want to take on some new theory, such as Gillett’s or a (possibly) modified version of Craver’s mechanistic explanation, that does cover those cases.
Not everything that scientists call a “law” describes a realization relation. Nor does every mathematical equation found in science describe a realization relation. The discussion will return to this topic.
Note that, in this case, some of the properties that realize the resistance are properties of the wire and not properties of parts of the wire. In the typical application of the Dimensioned view, it is the parts of s that have properties that realize the property G. This observation does not affect the issue of MRCD.
Many of the examples that will follow involve lengths. If one embraces the causal theory of properties that is part of the Dimensioned view of realization, then one might object that quantities such as length, area, and volume do not pick out genuine properties on the ground that length, area, and volume are not powerful properties.
There are many options for replies here. First, one might abandon the causal theory of properties and try to develop theories of realization and multiple realization that do not depend on it. Second, one might keep the causal theory of properties and argue that length, area, and volume are, in fact, powerful properties. Third, one might keep the causal theory of properties and argue that it is a sine qua non of the viability of the causal theory of properties for illuminating scientific theorizing that it correctly reconstruct the numerous scientific laws invoking length, area, volume, and so forth. In the case of resistance, such a reconstruction might go as follows. One might argue that it is not the length of the wire per se that, in part, determines the resistance of the wire, but instead the number of atoms in the wire and the way in which those atoms are bonded together that determines the resistance. It is, as the Dimensioned view of realization would have it, that the resistance of the wire is determined by the properties and relations of its parts.
The matter of sorting through these options is perhaps best construed as another respect in which these apparent examples of MRCD bear further scrutiny. Nevertheless, they are tangential to the present focus on MRCD. The causal theory of properties is prima facie not essential to the analysis of MRCD.
Notice that, in this example, one of the realizers of an action potential velocity is a property of the axon (i.e. the axon diameter), but other realizers are properties of parts of the axon (i.e., the membrane capacitance and the membrane resistance are properties of the axonal membrane, while the internal resistance of the axoplasm is a property of the axoplasm). In this example, compensatory differences are possible among the properties of the parts of the axon, such as the membrane capacitance and the internal resistance of the axoplasm.
An intriguing feature of this example is that, if psychological properties are realized by the velocities of action potentials, and if the velocities of action potentials are multiply realized by different combinations of membrane capacitance values, axon diameters, membrane resistance values, and internal resistance values, and if realization is a transitive relation, then one has some reason to believe that all such psychological properties are multiply realized by different combinations of membrane capacitance values, etc. In short, psychological properties might be multiply realized by cellular and subcellular properties.
An important ancillary feature of this case is that it suggests that there are cases of multiple realization of a single property instance over time. Let individual s bear the property G of having emmetropic or “20/20” vision. Insofar as changes to the radii of curvature of the front and back surfaces of the lens over time are precisely offset by changes in the refractive index of the internal components of the lens, one will have a single case of emmetropic vision multiply realized by different combinations of lens surface curvature and refractive index. Such cases are relevant to the debate between Gillett and Polger and Shapiro regarding the multiple realization of property instances. Cf. Gillett 2002, 2003, 2007 and Polger and Shapiro 2008.
See the comparison of the normal green cone opsin with the so-called R2G3 variant described in Merbs and Nathans 1993.
One might add bond angle to this list. It must be added to explain the realization of the dipole moment of many more complex molecules, but it is always the same in diatomic molecules, so discussions of these molecules will sometimes/often suppress this detail. In whatever way this is resolved it will be orthogonal to the existence and character of MRCD.
One reviewer provided the following comment:
It is really difficult to see how MRCD could work in the cases of psychological states. In the physical or neurobiological examples it is clear that there are measurable quantities that offset changes in each other so that the overall quantity (such as resistance) stays exactly the same. But how would this work in the case of psychological states, which are not quantitative? What would the compensatory differences look like in a non-quantitative context?
The same problem arises in the context of the Bechtel & Mundale issue (p. 17) and the discussion of Shapiro’s pain example (p. 20). The problem is that the author has not presented any concrete examples of genuinely psychological properties that would be MRCD, and it is hard to imagine what such examples would look like. This question should be briefly addressed at some point in the paper.
These comments raise a number of issues that cannot be fully addressed here. So, what follows is a start.
First of all, the principal point against Shagrir, Bechtel and Mundale, and Heil and Shapiro, is that they have overlooked a possibility. In the physical and biological cases, we have very solid examples of theories that are committed to MRCD as a nomological possibility. This suggests that for all we know, there are psychological cases like this. The physical and biological cases point out an epistemic possibility not considered.
Second, the reviewer invokes an implausible assumption. Having the capacity to visually resolve detail, being dark adapted to some degree, or having a particular cognitive processing speed are “quantitative” psychological properties.
Third, the reviewer’s suggestion that one cannot get MRCD for “non-quantitative cases” is unsupported. The reviewer gives no reason why there cannot be multiple realization by compensatory differences in cases such as the following:
I) Take some very specific visual capacity, such as the capacity to name Halle Berry upon seeing a particular image of her. Normal individuals might do this in a limited number of visual fixations, where individuals with a severed corpus callosum might require additional compensatory head and eye motions to ensure that a sufficient amount of information is available to both hemispheres to enable the reporting.
II) Take the capacity to assign a particular parse to the sentence, “Visiting relatives can be boring”. One way to do this might be to rely on semantic information regarding the subject matter under discussion; another might be to have a syntactic engine that defaults to one parse over the other. When an agent cannot rely on semantic information, the agent compensates by relying on the syntactic parser default.
Such psychological examples, however, bring us back to the issue Shagrir, Bechtel and Mundale, and Heil and Shapiro were exploiting. In the psychological cases, the lack of well-established theories of exactly what the realizers of psychological states are and how they work makes it seems plausible to doubt that the capacity to name Halle Berry or to produce the parse of “Visiting relatives can be boring” is exactly the same in both putative realizations. The strategy of this paper has been to try to establish the reality of MRCD in unproblematic scientific examples before trying to establish its reality in more problematic scientific examples.
One might speculate that what Bechtel and Mundale are claiming is that multiple or unique realization is a matter of how one chooses to describe properties. But, realization is presumably a relation among properties, not a relation between predicates or descriptions of properties.
The parenthetical qualification seems to be in order for the following reason. Shapiro 2008, pp. 522, considers an analog watch and a digital watch that have the same timekeeping properties, say, keeping perfect time. This seems to be a case of a property that is multiply realized (perhaps by two distinct sets of properties). Indeed, Shapiro has long allowed that analog and digital watches do multiply realize some properties. So, Shapiro does not uniformly reject multiple realization.
The cases of the scientific laws indicate that there is a nomological possibility of a single property being realized by distinct sets of realizers. One might, however, defend the Heil and Shapiro line in the following manner. One might first observe that Heil and Shapiro do not address nomological possibility, only what is likely or plausible in the actual world. Second, one might note that something can be nomologically possible yet not actually empirically likely or plausible; something’s being nomologically possible does not mean that it is actually a plausible occurrence. (This was noted by an anonymous reviewer for the journal.) This conceptual point is fair enough, but the purpose of this paper is to draw attention to MRCD, rather than “refute” Heil and Shapiro. It is consistent with this goal that sorting out the epistemological issues of the likelihood or plausibility of MRCD be left for another occasion.
The discussion here mirrors that found in (Aizawa and Gillett, 2011).
Some might prefer to mark the distinction described here by saying that two chemically identical photopigment molecules provide two tokens of the same type of realization of a given absorption spectrum, where two chemically distinct photopigment molecules provide two distinct types of realization of that absorption spectrum.
See Craver 2004, p. 967.
(NDR***) appears to endorse what was earlier described as the “eliminate and split” strategy for blocking multiple realization.
References
Aizawa, K., and Gillett, C. (2009a). Levels and Individual Variation: Implications for the Multiple Realization of Psychological In Bickle, J. (Ed.) Oxford Handbook of Philosophy and Neuroscience. (pp. 539–581).
Aizawa, K., and Gillett, C. (2009b). The (multiple) realization of psychological and other properties in the sciences. Mind & Language, 24, 181–208.
Aizawa, K., and Gillett, C. (2011). "The autonomy of psychology in the age of neuroscience." In Illari, P.M., Russo, F., and Williamson, J. (eds.) Causality in the Sciences. Oxford University Press. (pp. 202–23).
Bechtel, W., & Mundale, J. (1999). Revisiting multiple realization. Philosophy of Science, 66, 175–205.
Craver, C. (2004). Dissociable realization and kind splitting. Philosophy of Science, 71, 960–971.
Craver, C. (2007). Explaining the brain: Mechanisms and the mosaic unity of neuroscience. Oxford University Press.
Esfeld, M., & Sachse, C. (2007). Theory reduction by means of functional sub-types. International Studies in the Philosophy of Science, 21, 1–17.
Gillett, C. (2002). The dimensions of realization: a critique of the standard view. Analysis, 44, 316–323.
Gillett, C. (2003). The metaphysics of realization, multiple realizability, and the special sciences. Journal of Philosophy, 100, 591–603.
Gillett, C. (2007). Understanding the new reductionism: the metaphysics of science and compositional reduction. Journal of Philosophy, 104, 193–216.
Gillett, C. (unpublished). Making sense of levels in the sciences.
Grosvenor, T. (1987). Reduction in axial length with age: an emmetropizing mechanism for the adult eye? American Journal of Optometry and Physiological Optics, 64, 657–663.
Heil, J. (2003). Multiply realized properties. In S. Walter & H.-D. Heckmann (Eds.), Physicalism and mental causation (pp. 11–30). Charlottesville: Imprint Academic.
Koretz, J. F., Kaufman, P. L., Neider, M. W., & Goeckner, P. A. (1989). Accommodation and presbyopia in the human eye. 1: evaluation of in vivo measurement techniques. Applied Optics, 28, 1097–1102.
Merbs, S., & Nathans, J. (1993). Absorption spectra of the hybrid pigments responsible for anomalous color vision. Science, 258, 464–466.
Moffat, B. A., Atchison, D. A., & Pope, J. M. (2002). Age-related changes in refractive index distribution and power of the human lens as measured by magnetic resonance micro-imaging in vitro. Vision Research, 42, 1683–1693.
Ooi, C. S., & Grosvenor, T. (1995). Mechanisms of emmetropization in the aging eye. Optometry and Vision Science, 72, 60–66.
Polger, T., & Shapiro, L. (2008). Understanding the dimensions of realization. Journal of Philosophy, 105, 213–222.
Shagrir, O. (1998). Multiple realization, computation and the taxonomy of psychological states. Synthese, 114, 445–461.
Shapiro, L. A. (2008). How to test for multiple realization. Philosophy of Science, 75, 514–525.
Soom, P., Sachse, C., & Esfeld, M. (2010). Functional sub-types. Journal of Consciousness Studies, 17, 7–26.
Wimsatt, W. (2007). Re-engineering philosophy for limited beings: piecewise approximations to reality. Cambridge: Harvard University Press.
Acknowledgments
Thanks to Guillome Chappuis, Carl Gillett, and Larry Shapiro for comments on an earlier draft of this paper. Thanks as well to participants in the Computation, Realization and the Brain Workshop organized by Eli Dresner and Oron Shagrir at the Institute for Advanced Studies at the Hebrew University, May 15-20, 2011, and to the participants in the workshop on Multiple Realization in the Special Sciences, Department of Philosophy, University of Lausanne, Switzerland, May 11, 2011. Thanks, finally, to two anonymous reviewers of this journal for extensive comments on earlier versions of this paper.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Aizawa, K. Multiple realization by compensatory differences. Euro Jnl Phil Sci 3, 69–86 (2013). https://doi.org/10.1007/s13194-012-0058-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s13194-012-0058-6