Is holism likely incompatible with scientific method?

Question

I've been reading on the holism-reductionism continuum. Reductionism strikes me as compliant with general principles of the scientific approach and method because it seeks to explain complex phenomena by breaking them down into smaller units and building theories bottom up, focusing on specificity and detail. Holism is the opposite and seems to me it could easily serve as a seemingly respectable (who doesn't want to tie it all together in a big "theory of everything"?) curtain for poor, insufficient or downright misleading and sloppy research. This is not to say we shouldn't seek to explain relations among phenomena -- only that it should be preceded by a thorough understanding of individual parts.

In modern scientific circles, is holism generally regarded as a conduit for pseudoscience?

Answer 1

I think modern scientific circles regard just about anything as a conduit for pseudoscience these days. However, I don't think holism is given any special treatment.

If you consider the ultra-pure-ivory-tower version of science, where you develop a hypothesis about the tiniest facet of the universe and then spend an infinite amount of time testing that hypothesis, then everything can be thought of as a "ground up reductionist effort" built on those tiny proven hypotheses. Real world science, however, has never been so static. There are always ideas flittering about, not yet nailed down into a bottom-up structure of theories. Quite often these ideas arise by observing gestalt behavior. Once observing these things, science then does strive to reduce it, for it is far easier to develop a scientific experiment in accordance with the scientific method if you take a reductionist approach.

However, they are very much dependent on these ideas that haven't been pinned down. They point the scientists where to go next. We may not see it much in physics, where atomic theory rules the day, but in biology there's myriad gestalt effects to consider. Over time, we try to reduce them, but there's no proof that all things can be reduced this way. Even in physics, Atomic Theory is taken as an axiom at most, and a guiding theory at least.

Consciousness is one of those things. While science may try to apply reductionism to it, so far it has not succeeded. Thus, any science which studies consciousness must embrace some level of holism because they do not yet have the reductionist structure to build upon.

Like all things, it's a balance. If science falls too far towards reductionism, it finds itself sluggish in responding to newly discovered gestalt patterns. If it falls too far towards holism, it finds it difficult to create repeatable and reproducible scientific experiments.

Answer 2

The issue is (or should be) a perfectly practical one. There are systems in the world whose effects we can experience and wish to understand but for which the cause of those effects is not completely known. Pure reductionism would slow science to a standstill for complex systems (like the weather), as developing theories that predict their outcomes based on an understanding of their component factors would take centuries of work.

Fortunately we already have scientific means of dealing with complex systems holistically without resorting to "quackery". The primary tool is the use of probability to denote those unknown factors and the extent of their influence. This works perfectly well for things like weather systems, epidemiology, genetics etc.

Friedrich Hayek first proposed after Popper that complex systems could not (and should not) be approached by the same reductionist science as simple physical systems, but that rigour could be maintained by the use of models with probability predictions.

For example, Stuart Kauffman applied statistical modelling of random boolean networks to show that attractors existed in gene regulatory networks without the need to demonstrate each individual interaction (which was not and still is not fully known). The work of Bill McKelvey takes Kauffmans' ideas and adds some further mathematical rigour, but still without reduction (if anything he adds further holism). You may be interested to read his paper www.billmckelvey.org/documents/Thwarting%20Faddism.pdf which specifically addresses the issue of "fadism" in complexity science.

Answer 3

I'm going to discuss this as though it were just a practical problem, when in fact it is a real problem that points to the boundaries of what science can do. Suppose you have a holistic theory where "everything" matters; how are you going to test it? how is it going to make predictions? In this way science, as an endeavor geared towards making testable predictive theories cannot admit too much holism. (To some extend, that is why various forms of pseudoscience incorporate holism: by allowing everything to affect everything else proponents can brush aside apparent contradictions as being caused by one of the uncontrolled contributors).

There are areas in science where more limited forms of holism, of a sort, are used. Population genetics (which can, AFAIK, be studied without detailed knowledge of the structure of DNA), some forms of material science (e.g. formulating the dynamics of granular material at a bulk without going into the details of their interaction) etc. And these are interesting/valuable areas of research.