Entailment

Excerpted from "Life, Itself"; page 57:

Robert Rosen wrote: "As philosophers have pointed out for millennia, all we perceive directly are our selves, together with sensations and impressions that we normally interpret as coming from "outside" (i.e., from the ambience), and that we merely impute, as properties and predicates, to things in that ambience. The things themselves, the noumena, as Kant calls them, are inherently unknowable except through the perceptions the elicit in us; what we perceive as phenomena, which are to an equally unknowable extent corrupted by our perceptual apparatus itself (which of course also sits partly in the ambience).

We can simplify things somewhat if we ask the more restricted question: is there any kind of entailment at the level of phenomena? Or, stated otherwise: does it appear to us that a phenomenon can entail another? The problem is still difficult, because entailment at this level is a relation between phenomena (just as inferential entailment is a relation between propositions), and we usually do not directly perceive relations. Indeed, a relation between phenomena depends on a double imputation: the first from sensation to phenomena, the second from phenomena to relations between them, Thus, if our knowledge of phenomena is already once removed from the ambience, any talk of entailment, or any other kind of relation between phenomena, is twice removed. On top of all of this is a further problem, that what we perceive is only a sample of what we could perceive and the problems of induction arising therefrom; see section 2C above.

Nevertheless, it is hard to believe, for instance, that we could use natural language, in its semantic role of bringing external referents inside, if there were not a great many phenomenal entailments; semantic language by its very nature impute hordes of entailments to the ambience, without going really dramatically astray. For this, and similar (albeit subjective) reasons, we will suppose that relations of entailment do indeed exist between phenomena; the question then becomes not whether, but when, such relations hold.

It was, of course, Aristotle who associated the notion of entailment between phenomena with the question "why?" and answered it with a "because". Indeed, the pair consisting of the question "why A?" and the answer "because B" precisely asserts an entailment of A by B, and hence, an explanation of B in terms of A. In this way, entailment relations between phenomena are subsumed under the general framework of causality. To the extent that science is the study of entailment relations between phenomena, Aristotle correctly identified science with the study of "the why of things" and scientific explanation with the elucidation of causal sequences.

We shall thus accept this view, that entailment relations can exist between phenomena and that their study comprises causality; hence science and causality are to that extent synonymous."

Science began as a study of causality; but what we really need to know about is causal entailment. Because, in order to understand causality, science must learn about the underlying entailments. The underlying entailments are a much larger capability than the causal phenomena they entail at any given moment. Causality is what we observe happening and, therefore, is what science must use to get AT the entailments underneath. That's our only avenue into the underlying entailment patterns, which can teach us about Natural Law. One of the aspects of Natural Law, in RR's view, is that entailment relations are consistent, regardless of the venue. I've tried to explain this before and it's a tricky concept to put into words. Why is it that we can build models that actually predict behavior in the system being modeled (if we do it well enough, that is)? Behavior that hasn't happened is not causality, right? What we're dealing with is actually entailment.

The trouble with focusing too much on "causality" is that there are many explanations which can seem to account for any given phenomenon. The phenomenon itself is just an "effect". Sometimes science attempts to replicate the effect, in the belief that if they succeed (simultion) they will have the causal entailments as well. But it doesn't work that way in a complex universe like ours. Furthermore, it's really not good enough to settle for a plausible explanation for causality if the explanation (model) only seems to commute in one kind of system but not another.... That's the situation science is currently IN (physics and the machine metaphor). Robert Rosen's assessment of the situation was that we don't have the entailments quite right and "close enough" to work in simple systems but not in complex systems is not acceptable. So ultimately, if we want to understand "causality"-- meaning any effect, any phenomenon-- we need to learn about entailment patterns.

Judith