¹ I shall not, as for many purposes one must, distinguish Marx's views from “Marxism” in this paper.

² Indeed, his views seem to entail that man is in such a way free, but I shall not argue that in this paper.

³ See his The Materialist Conception of History.

⁴ See his The Role of the Individual in History.

⁵ One could just as well state the doctrine in terms of a parallelism between mental states on the one hand and either behavioral or brain states on the other rather than between mental and neurophysiological states. Probably in truth the “strictest” parallelism is to be found between brain states and mental states.

⁶ For that reason, one may want to speak of bodily states as the cause of mental states and not vice versa. This is my cue for another comment. When there are parallel mental states, one can substitute them for their parallel neurophysiological states in particular explanations. Since however there are a number of different neurophysiological states which have no parallel mental states, one cannot always do this, as one could in our model universe. For this reason I shall speak as if only the neurophysiological states interact with other physical variables, i.e., that the mental variables are “merely*' parallel and not ”interacting.“ This way of speaking is also useful in distinguishing psychophysical parallelism from what is usually called interactionism. The latter would be the view that the mental variables interact with physical variables and that those mental variables are not tied to any physical variables by parallelistic laws. In short, this is the doctrine that the physical variables do not constitute a causally closed system. Such a view, it may be worth noticing, is perfectly compatible with determinism.

⁷ This kind of “materialism” is also Freud's. Many Freudians make the mistake of taking Marx to be a materialist in the sense of denying the existence of consciousness, while many Marxists make the mistake of taking Freud to deny the physical “basis” of consciousness.

⁸ I of course include the relations in which a thing stands as part of its properties. Occasionally I shall remind ourselves of that.

⁹ Briefly, to say that a theory is comprehensive (of its area) means (1) that all the property terms of the area either are “in” the propositions of the theory, defined or undefined, or that they are definable in terms of those property terms “in” the propositions of the theory; and (2) that all the laws of the area, whether known or not, are deductive consequences of the propositions that make up the given theory. For more details on this and other issues of reduction, see Gustav Bergmann's “Reduction” which appears in Current Trends in Psychology and the Behavioral Sciences, University of Pittsburgh Press, 1954.

¹⁰ This doesn't mean that we may not want or need definitions of these property terms. But their definitions will be in terms of other properties of the same “level” or science. The importance of that will become obvious in a moment.

¹¹ By sociology I shall mean all the social sciences considered as one vast science. With respect to the issue of freedom, we are interested in the relation of certain group variables to the behavior of humans, of course. So I shall speak mainly of the connection between sociological or group variables on the one hand and psychological or individual variables on the other. By a group variable I mean one which is plausibly understood in terms of humans and their behavior. Or perhaps one could say that a group has humans as its constituents as long as that is not taken to beg any questions about the “connection” between the group and its constituents.

¹² Some sociological variables such as the anger of the crowd would reasonably be defined in terms of the properties of (and relations among) humans alone; others such as the condition of the railroads would reasonably be defined in terms of the properties of individual but non-human things alone; still others would reasonably include the properties of both human and non-human things in their definitions. A plausible example is the total social product.

¹³ That is, if the group or state exemplifies some property which cannot be understood in terms of the properties of “its” constituents, then it would be natural to say that the group or state is more than its constituent individuals. “The whole is more than the sum of its parts.”

¹⁴ Composition laws are in principle dispensable anyway, as can be seen from my discussion. The reason I introduce the discussion is because some have mistakenly believed that if the behavior in the more complex situation is not computable in some way from that in the less complex, then determinism is false.

¹⁵ That is, such that it varies with some variations in the individual's behavior. There may be some disjunction; that is, a group variable may be defined as microstate A or B or C or... and so on. The commonsense idea is that we can effect a change in many ways; but there must be a limit. If the change occurs no matter what we do, then that limit has been far surpassed.

¹⁶ One may want to distinguish what Bergmann calls historical from systematic process theories. The former has laws of the sort: If the system is now in state x and was earlier in state y, then at such and such future time it will be in state z. The italicized phrase indicates the difference; that is, in a systematic process theory the laws would allow one to compute any state of the system from any other single state. It may be useful to understand Marx as looking for an historical rather than a systematic process theory of society. See Bergmann's “Reduction” referred to in footnote 9.

¹⁷ This is reflected in the repeated claim of some Marxists that history is independent of man's will. One thing this could mean is that the states of society are computable given the values of the social variables alone, i.e., that there is process at that level. But we have just seen that that does not imply that man isn't free in our sense.

¹⁸ Brodbeck, May, “Methodological Individualism*:: Definition and Reduction,” Philosophy of Science, Vol. 25, No. 1, January, 1958, p. 21. Several of the points discussed in my paper are covered in more detail in hers. I have also made use of her terminology at many points.

¹⁹ Or we can treat that as the value of some variable. The distinction between a variable and its values depends on the context. Being red and being blue are values of the variable being colored; but being pink and being scarlet are values of the variable being red.

²⁰ Such an ideal is in a stronger sense unobtainable if minds interact in the sense of ‘interactionism’ explained in footnote 6. For one could never determine a state description for a given moment if (1) he could not observe mental states and (2) those mental states had no parallel brain or neuro-physiological or behavioral states. That there is some kind of parallelism between what a person thinks and what he does is of course commonsensical.

²¹ Nor, if they are macrovariables, do the properties in terms of which they are defined interact with the non-economic. Of course, Marx wanted to say that in a certain sense almost all of man's behavior is “economic” under capitalism. But the argument I am considering makes sense only if we ignore this.

²² They will “make a difference” to simultaneous events, so to speak, if the group variables are to be defined in terms of individuals' behaviors. But on the view under consideration those social variables are non-interacting.

²³ Letter of Engels to Franz Mehring dated London, July 14, 1893, as it appears in Marx and Engels, edited by Lewis Feuer, Anchor Books, 1959, pp. 407-408.

²⁴ Letter of Engels to Joseph Bloch dated London, September 21-22, 1890, ibid., 397-398.

²⁵ To say that they are not process laws means here that if an event of the kind mentioned in the antecedent of the law occurred and if certain other events occurred which as a matter of fact do not, then the corresponding event of the kind mentioned in the consequent of the law would not occur. That is, it may be that “there is” a cure for cancer but a matter of fact that no one will ever discover it. So, if certain choices were made, Marx might say, then communism would not succeed capitalism, but as a matter of fact those choices won't be made. See Section IV.

²⁶ Or, if you like, that either (1) the group variables which are defined in terms of individual behaviors are interacting variables or (2) there are non-trivial parallelistic laws between neurophysiological states or behaviors and the values of the group variables.

²⁷ Letter of Engels to Joseph Bloch dated London, September 21-22, 1890, as it appears in Marx and Engels, p. 399. In some of his more speculative moments Marx suggested that under communism for the first time the course of events would be as most men willed. This is probably what Engels was referring to when he spoke of “the ascent of man from the kingdom of necessity to the kingdom of freedom.” But that is another story.