These papers helped to establish determinism (or compatibilism) as the dominant view of philosophers in the late twentieth century, at least until Peter van Inwagen's 1982 Essay on Free Will restated the standard arguments against free will as his "Consequence Argument" and "Mind Argument, and Robert Kane's 1985 Free Will and Values claimed quantum indeterminism could provide human freedom without jeopardizing control."

In the introduction to Free Will and Determinism, Berofsky laments the lack of a good definition of determinism in most works on human freedom.

In discussions of human freedom it is not uncommon to omit a definition or clarification of the thesis of determinism, although reference to it may be made. This is quite serious if one considers (1) the fact that this thesis often plays a fundamental role in conceptions of human freedom and (2) the possibility that resolutions of fundamental questions about human freedom hinge upon clarification of the thesis of determinism.

In 1971 Berofsky published his book Determinism, in which he gave an abstract and formal definition of determinism in mathematical, logical, and analytical philosophical terms:

(x) [x is an R-sentence ⊃ (∃y) (∃z) (y is a state description-sentence • z is a law-sentence • (y • z) ⊃ x)]

An R-sentence is true, contingent, singular, and temporal. It must not contain a causal, evaluative, or basic dispositional predicate. Any logical consequence of either an R-sentence or any conjunction of R-sentences is an R-sentence. If the predicate of the R-sentence is a determinable whose determinates admit as values any or almost any real number, the laws in the deterministic account of this sentence must contain variables for that determinable that may also take values from any or almost any real number (Continuity-requirement).

A state description-sentence is a conjunction of any mathematical, logical, and linguistic truths, any true sentences of the form "x = y," and the state description proper. The state description proper is a subset of R-sentences except that sentences containing basic dispositional predicates are allowed, and sentences containing positional and temporal predicates are ruled out. The state description must be logically and criterially independent of the sentence the deterministic account is about; and no sentence may refer to a time period subsequent to t where t is the temporal designation in the sentence the deterministic account is about.

No other temporal restrictions on state descriptions are made.

A law (as used in the definition of determinism) is a true, logically contingent universal conditional with non-self-contradictory, open-ended, nonpositional, and qualitative predicates that meets five requirements (DS, C, S, AE, D). ("A law" may actually be a conjunction of laws.)

A deterministic account of some R-sentence is, of course, a specific state description-sentence and a specific law (or laws) that entails the R-sentence. Thus, any deterministic account of x is a deterministic account of each logical consequence of x. Moreover, a deterministic account of an R-sentence may be povided by several deterministic accounts of sentences which conjunctively entail that R-sentence.
(Determinism, pp.268-9)

1. Determinism is a claim about the world that is true or false,
   insofar as determinism is a proposition.

   The libertarian believes that humans are not free if determinism is true. The reconciliationist (compatibilist), accepts determinism and believes that it does not rule out freedom.

2. Determinism is not true in virtue of linguistic or other conventions alone.

   The truth of determinism is guaranteed by virtue of its definitions and certain scientific conventions. Indeterminism is virtually self-contradictory, with "devastating consequences" for the free will debates.

3. Determinism is prima facie incompatible with human freedom

   In terms of the notion of law, all human behavior is law-governed and thus not free. But Berofsky asks whether laws are necessary, and concludes that they are not. This reduces the truth staus of determinism to three: contingent truth, contingent falsity, and necessary falsity (p.291). He concludes that the scope of determinism is limited. Since human actions cannot be determined, he considers the possibility that determinism is necessarily false. In Information Philosophy, nothing about the physical world is necessarily true or false.

Berofsky cites what he calls a "univocity condition," that the meaning of the term "determined" in "All events are determined" has the same meaning in the statement "This event is determined." But obviously, some events may be determined and others not, since some causes may not themselves have causes, which philosophers since Aristotle have noted.

Berofsky also defines a "unicity condition," that what occurs in the world at any time t restricts the world to just one possible future. This is the view of Laplace and the concern of William James in The Dilemma of Determinism.

Berofsky says that any definition of determinism must capture this idea, but we prefer to call this idea "predeterminism" to contrast it with the adequate determinism we have in the real world, which admits some indeterminism but keeps it under control, limiting chance to the creation of alternative possibilities for action.

Berofsky considers other definiens for determinism, including human and divine foreknowledge (p.9), predictability, which he claims may not require foreknowledge (p.28), and universal causality, the doctrine that every event has a cause.

He compares a generality theory of causation to a regularity theory (which may not be "necessary"), a thought which he associates with David Hume (p.42)

Berofsky is a strong supporter of Hume. He says that "no great philosopher has been treated as unfairly as has Hume by philosophers on both sides of the Atlantic." (p.6)

In 1987, Berofsky claimed in his book Freedom from Necessity that

No philosophical problem is more deserving of the title 'the free will problem' than that concerning the assessment of the claim that a deterministic world, i.e., one governed completely by universal scientific law, has no room for agents with dignity sufficient to justify attributions of moral responsibility to them.

We shall also extend the term "incompatibilist" in a perfectly natural way to one who applies his reasoning to a particular case by concluding that an agent is not morally responsible for an action in the event that that particular action is determined.

It is a psychological mystery to me why so few writers on this subject note the significance of this extension. Although it is virtually a truism that an incompatibilist draws an important conclusion about some determined action regardless of the truth or falsity of the general thesis of determinism, many philosophers regard the free will problem as having been deprived of its motivation once determinism is surrendered. This response is as viable as a refusal to see the danger in an arsenic-laced apple because not all the apples in the basket are laced with arsenic.

In Berofsky's last book, Liberation from Self in 1995, he turns to psychology to explore the notion of human autonomy. Defying the etymological origins¹, he denies that autonomous agents are individuals whose actions are self directed.

1. αὐτονομος, ον, living under one’s own laws, independent, of persons and states; generally, of one’s own free will.
Liddell, Scott, Jones: A Greek-English Lexicon. Oxford University Press, 1996, S. 281

I will argue that autonomy is essentially constituted by the manner in which an agent is engaged in her world rather than the metaphysical origin of her motivations. (p.1)

Critics of the view that freedom is the liberty of indifference are fond of observing that our freedom is not threatened by the need to subordinate ourselves to the rules of mathematics and logic. We do not feel that we would have greater freedom (at least of a worthwhile sort) if we retained the ability to withhold judgment in the face of deductively conclusive evidence. I believe that we should feel similarly unafraid by the correlative need to subordinate ourselves to the other areas of life in which we act and find our satisfactions. In this way, autonomy is achieved. (p.2)

...there is something perverse in wanting the freedom to reject a theorem one knows to be true. (p.2)

Autonomous persons must be independent, not just from the pernicious influence of others, but from the pernicious influence of their own earlier lives. For that robust engagement with the world is possible only for persons who have liberated themselves from the disabling effects of physiological and psychological afflictions. We are all limited and, perhaps, our lives are completely determined. But there are crucial differences in the manner in which our earlier life bears on our later life. The possibly deterministic process that has brought us to our current state may have an independence and authenticity depending on the character of current interactions. The autonomy of a sculptor is grounded in knowledge of the craft, its techniques, its history, its standards, in his openness, and in his ability to respond to relevant inputs, his work as it has so far evolved, his own responses to that work, and relevant features of the world such as the manner of his involvement in professional life. If these elements are in place, we do not require in addition a radical rupture with the past.

On the other hand, when we have formed and retained an irrational belief that is rendered immune from learning through experience, when we respond to the world in a cognitive style that is prone to produce distortion, when our experience is filtered through subjective principles designed to reinforce prejudices which serve a defensive function, we have failed to transcend our origins in a way that bears on our capacity to establish an autonomous connection to the world.

In order to characterize impediments to autonomy, therefore, we must turn to psychology for a deeper understanding of the ways these barriers can form. It is striking to observe that this is done by very few philosophers who are interested in autonomy. (p.2-3)

In 2003, Berofsky contributed an article on "Classical Compatibilism" to Moral Responsibility and Alternative Possibilities, edited by Michael McKenna and David Widerker.

He describes a "Strong Compatibilism" that rules out alternative possibilities and claims that "Classical Compatibilism" does not rule them out:

Those philosophers who draw depressing consequences about human freedom and responsibility from the existence of deterministic accounts' usually do so because they draw an intermediate conclusion that has been characterized in a variety of ways: from the fact that a particular action is determined, the conclusion is drawn that the agent could not have done otherwise (or that the action is necessitated or that there is no alternative possibility or that the action is inevitable, and so on).

Some compatibilists believe that freedom does not require determinism and, therefore, concede to the incompatibilist the possibility of a freedom-assuring indeterministic process of deliberation leading to a decision D at t1. Some of these compatibilists believe as well that such a process can exist alongside a distinct mechanism that insures but does not determine that D occur at t1. The presence of the distinct insuring mechanism guarantees that the agent could not have done otherwise, while the agent's freedom over the decision is ostensibly assured by the fact that it is the result, not of the distinct mechanism, but rather of the deliberative process. The latter incorporates whatever elements are deemed essential for freedom, one of which the libertarian believes is indetermination. The incompatibilist's complaint that determinism rules out freedom and moral responsibility by ruling out an agent's ability to do otherwise is thereby rendered otiose — determinism is harmless even if it entails that everything we do we must do. Call this strategy 'strong compatibilism,' a label whose qualifier designates immunity from refutation by the 'discovery' that deterministic accounts entail that an agent could not have done otherwise.

Strong compatibilists do not believe that all freedom-assuring deliberative processes are indeterministic. They are compatibilists, after all, and believe that the key freedom-conferring elements — deciding not under coercion or compulsion for reasons one is neither constrained nor compelled to possess — may be found in some deterministic processes as well. In this respect, of course, they part ways with the libertarian. Their position is stronger than classical compatibilism, for the latter accepts the incompatibilist demand for alternative possibilities, but supposes that the demand fails to be met only when specific sorts of determining conditions are present. Classical compatibilists agree then with incompatibilists that the presence of a distinct process that renders the agent unable to choose otherwise negates the agent's prima facie freedom conferred by the deliberative process. They both see the agent as the victim of an illusion. For he deliberates in the belief that he faces open possibilities. It is also of some interest that, in the eyes of the classical compatibilist, since determination per se neither destroys freedom nor renders an agent unable to do otherwise, the destructive elements of the process that insures the decision must be different from simple determination.

In recent journal articles, Berofsky has challenged the idea of "source incompatibilism", whether Robert Kane’s event-causal libertarianism or the various agent-causal varieties defended by Derk Pereboom and Randolph Clarke. He says

We are very different from machines and the lower animals. We are complex creatures who often deliberate about a multiplicity of options. We reflect and make decisions, often rationally, that we do not have to make. We occasionally have original ideas and create original products. We are to some extent responsible for the lives we lead, including the decisions we make...Even if we find ourselves with inculcated values, we can sometimes alter them upon reflection and adopt a new set of values.
(Philosophical Studies 131 (2) p.419)

Chapter 1.4 - The Philosophy	Chapter 1.6 - The Scientists
Home	Part Two - Knowledge