So, one obvious way of responding to problems with a conception of dialectics as the "laws of motion" of all reality is to re-interpret dialectics as the "laws of motion" of history, or just of capitalism. Then we might reject Engels' "Dialectics of Nature", but stop there.
This should be a fundamentally unappealing solution to a materialist, though. It assumes an untenable absolute division between nature and society. Raymond Williams quotes Marx: "One basis for life and another for science is a priori a falsehood."
Even if we wanted to violate that stricture, where would we draw the line? As Bertell Ollman points out, "People have bodies as well as minds and social roles... capital, commodities, money, and the forces of production all have material as well as social aspects." How could we analyze capital and its organic composition without speaking of machinery and the development of technology, or labor and its alienation without speaking of human bodies and their limits?
There is another problem with isolating dialectics as a science of human history. It is that we do, in fact, sometimes see examples in physics, chemistry, and biology which involve no necessary human intervention but nevertheless obey certain dialectical "laws"; examples where contradictory tendencies cause change, where the accumulation of quantitative change leads to a qualitative state transition, etc. Many of Engels' examples in nature - evaporation, Darwinian evolution, the life cycles of plants - do seem to match dialectical patterns, often more clearly than most social processes. If dialectics does not truly apply to the spheres in which we find these examples, what are we to make of them?
Ollman and John Rees offer essentially the same answer to the question of the scope of dialectics: that while the dialectic takes different forms when applied to humans and to non-human nature, it remains, fundamentally, universal. Rees suggests we say that "dialectical development [is] a feature of the natural world as well as the social world without... assert[ing] that the form of the dialectic [is] the same in both cases." Ollman asserts that "movements on each level of generality must be seen as expressions of laws that are specific to that level as well as versions of more general laws."
A first response to this is that if we have already rejected a conception of the dialectic as a set of laws governing all reality, we need to reject it as a set of laws governing history or capitalism, in order to preserve a unity of method corresponding to the unity of the material world. That's simple enough.
But there's something more suggested by Rees' reference to different "forms" of the dialectic, and Ollman's reference to different "levels of generality". What does it mean that the dialectic is the kind of thing that can not merely be expressed differently, but take a different shape itself, in different cases? That it is the kind of thing that can only be fully defined within a given system, and not as a law identically applicable to all cases? That - if earlier posts were right - there are real things, processes and relationships about which dialectics has nothing to say?
I think the correct conclusion to draw here - though neither Rees nor Ollman explicitly draws it, and Rees, at least, would deny it - is that "law" is the wrong concept. When we look at examples of dialectics in action, we do not really see reality obeying a common set of rules or logic. Rather, we see change taking a certain structure, a form which shares a sort of family resemblance with the forms we see change take in many other places. So perhaps the dialectic is not analogous to the theory of relativity, but to the notion of an unstable equilibrium or that of the positive feedback loop, recurring patterns that we see again and again in different fields.
A conception of the dialectic as a family of forms allows us to make sense of the "unity in difference" of human society and non-human nature: a common basic structure which may take different shapes, or be more or less common, depending on the process. And as above, I think it is hard to deny that we see many real phenomena which have some dialectical aspect, so-defined. But this still leaves open questions.
One set of questions revolves around why we see these forms - and why we should care. In rejecting the dialectic-as-law, we have rejected the idea that change is necessarily structured along these lines. But perhaps these forms are characteristic of capitalism? If so, why? Is there a worthwhile approach to inquiry that should lead us to pay special attention to these forms? Or is dialectics good for nothing more than constructing a rather eccentric and abstract catalogue?
Another set of questions starts from an even more skeptical position. Is there really anything substantial in common among all the forms that are grouped together under the name "dialectics"? If we see some dialectical aspects of a situation, does that give us the right to conclude anything further that we do not already know? Or does "the dialectic" not describe any single coherent thing - at least not if interpreted as a structure of change?
Future posts will take up these questions - starting with the latter set.
 Marxism and Literature, p. 63.
 Dance of the Dialectic, p. 70.
 Algebra of Revolution, p. 75.
 Dance of the Dialectic, p. 97.