Serendip is an independent site partnering with faculty at multiple colleges and universities around the world. Happy exploring!

You are here

Digital Determinism: Why It Is Worth Taking Wolfram Seriously

PaulGrobstein's picture
I have no interest in defending Wolfram as either a person or an academic scholar against the kinds of criticisms expressed (appropriately I think) in class today. I do though want to explain and justify my characterization of his work as "digital determinism" and as a unique/important "coherent and comprehensive explanation of everything". And hence as, whatever its shortcomings, a body of exploration/thinking that it is important to understand and pay attention to. I earlier argued that "computer models are not capable of nor aimed at determining what is 'real'" but instead are intended "to establish that some pattern/phenomenon that is presumed to depend on complexity/planning/a directive element can be produced without that. To show what might be, rather than what is." Wolfram's work needs to be appreciated in these terms. It is an assertion that one might in principle account for all known phenomena (literally ALL, from physics through biology, psychology, sociology, history, and, yes, art) in terms of very simple things (locations having only two possible states) interacting in very simple ways (locally and deterministically in digital time steps). The core of this argument rest on the possibility that the universe is itself digital in both space and time if looked at at a small enough scale, that each location is either occupied or not occupied at successive times by some fundamental particle, and that what has to be accounted for is the successive patterns (viewed at multiple different scales) formed by the occupied and non-occupied locations. What Wolfram is asserting is that any pattern (and succession of patterns) can in principle be produced by nothing more complicated than a set of rules of the sort exemplified by the one dimensional cellular automata he has explored. In fact, he is suggesting that one needn't even add the complexity of three dimensions or entertain the possibility of greater numbers of states (to say nothing of more complex patterns of interactions), that certain rule sets (eg rule set #110) operating on one dimensional arrays are capable of producing any conceivable complexity of pattern and pattern succession. That's why I characterized Wolfram's claim as "digital determinism"; there is no need to worry about mushy between 0 and 1 values nor to be concerned about "random" processes (to see nothing of "meaning" or "intentionality"). And that's why I referred to Wolfram's claim as a"unique/important 'coherent and comprehensive explanation of everything". And that's why I (for one) am not uncomfortable with his assertion that he is creating "a new kind of science". If Wolfram is right, the problem of science (ALL sciences) reduces to the problem of determining no more than two things: the rule set that is in fact used by the universe to go from one digitial state in digital time to the next and the starting configuration of the original digital array. That's indeed a new and quite general "science". Wolfram's conjecture unquestionably has antecedents and reflects explorations by others that are not acknowledged by him but it is, to my knowledge an unusually clearly stated "bold conjecture" that, for this reason, warrants serious attention by anyone with "scientific" interests/credentials. The challenge for people like me, who are skeptical of the claim, is to make clear, if possible, in what way the phenomena we wish to make sense of cannot even in principle be accounted for by "digital determinism" or, failing that at the moment, to make a coherent argument for why pursuing the program of "digital determinism" is less likely than pursuing some other agenda to uncover (or create) phenomena that cannot be accounted for by "digitial determinism". My intuition is that Wolfram's "new science" is incomplete, that there are phenomena created by analogue systems, and/or by systems that involve non-deterministic interactions, and/or by systems that involve non-local interactions, and/or by systems with more complex architectures (including "top down" influences) that cannnot be accounted for in the terms that Wolfram proposes. And that, precisely, is why Wolfram's work is worth paying attention to. The old intuition that complex outcomes require complex explanations no longer suffices. It is incumbent on those of whose who believe Wolfram's program is "too simple" to come up with ways to make that argument as clearly and compellingly as Wolfram has made his. Wolfram may be "wrong" (almost certainly is) but I have yet to hear the argument that establishes that convincingly by providing an equally clear, coherent less wrong perspective. I'll keep working on that problem, and invite others to join in.


Doug Blank's picture

Thanks for the comments, Paul. Just a brief note right now: I do think that whatever you want to call what Wolfram has done in his book, it really is as import as what Darwin did in his. If it seemed otherwise, I just wanted to show that we show look at all claims with some skepticism, but be prepared to accept them no matter how they are packaged, or who says them. We got a peek yesterday that science is not just a bland mechanical process, but there are real people "doing it" and each of them have personalities, egos, lives, etc. To me, if something "account[s] for all known phenomena", then there is something very "real" about it, whatever we mean by "real". I may believe that I am heading for the "truth"; you maybe "getting it less wrong". Either way, we still do the same kind of science every day. Finally, take a look at: More later...
PeterOMalley's picture

To me, one important consequence of Wolfram's theory is the "deterministic" aspect of it. Every outcome of cellular automata (and, by implication, the world) is that it is completely deterministic. The "head-heart" article had proposed (I believe) that the key to understanding brain function may lie in understanding emergent complexity as it applies to neurons and neural networks. If this sort of thing is deterministic, it doesn't allow for much sort of free will. I am inclined to agree with you, Paul, when you intuit that there are some phenomena that would require non-deterministic interactions or "top down" influences. Perhaps free will (if it exists) is some sort of "emergent" result of neural networks, together with "top down" influences? Well, this is pretty rampant speculation from me, at this point. I think it would be very interesting to look at this more in depth, though.