Science is such a problematic philosophical phenomenon because of its double nature: on the one hand, it has had some astonishing epistemological successes, i.e., it has given us some wonderful objective knowledge about the world, both natural and non-natural; but on the other, it is a heterogeneous social practice and as such it reflects the good and the bad of us humans. Thus, some philosophers have thought that science must be essentially the subject of epistemological reflection, while others think the most accurate portrayal cannot come but from a sociological, political and/or ethical approach. Science is not merely an epistemological enterprise, but also a political one.

Each one, unsurprisingly, tries to incorporate the insights of the other, while criticizing it as reductive. My own philosophy of science teacher, and now colleague, Ana Rosa Pérez Rasansz, likes to say that one of the central questions in the philosophy of science is not whether it is lead by reason or by more mundane social forces, but how can it be so rational despite being driven also by all sorts of social forces –– I am paraphrasing, of course, and her statement should not be interpreted as assuming an ahistorical notion of rationality, of course. All we need to recognize is that our notion of rationality is closely linked to that of scientific knowledge, so much that any account of science that dismissed it would be incomplete to say the least.

Even the so-called demarcation problem is going to be conceived differently in each tradition. Within the epistemological tradition, Popper’s tradition, the question is how do we distinguish between representations that deserve the epistemological prestige we ascribe to scientific theories and models. Instead, I think that the question must be: how the limits of what is counted as science is constructed.

Central to this debate is the challenge of accurately explaining the dynamics of scientific evolution—that is, of how science changes over time. It is evident that such evolution cannot be characterized merely as the accumulation of new discoveries, given that the historical record abounds with errors, significant misjudgments, prejudices presented as facts, and even instances of unethical and criminal conduct. Nevertheless, any viable model must still account for, or at least account for the illusion of, scientific progress, properly (and historically) understood.

Lakatos inaugurated a strategy that centrally involved two basic ideas: first, that the basic unit of historical change is not obvious; i.e., when we talk about scientific change, it is not immediately clear what it is that we are saying changes. Following Eckardt von B (2006), Atocha Aliseda reminds us that as science changes, it is not only the theories that change, but also our conception of what is being studied (its so-called “domain”) changes. The questions and problems it tries to answer and solve change. Its paradigmatic examples change. Its basic assumptions change, etc. Methods also change. But not only that, the criteria for who has the authority to participate in this or that scientific practice also changes. Its prestige changes, etc. Thus, for example, we cannot speak simply of changes in biology, but rather need to be more specific about which parts of this complex system we call “biology” are changing in a given historical episode. And what might seem like a normal accumulative change when we focus on a certain aspect of scientific practice, might seem like earth-shattering when we focus on another. That is why, as Lola Almendros, reminds us, some changes will seem more profound to us than others, and why we sometimes find it useful to talk about “revolutions” and “paradigm shifts”, even when continuities are abundant and conspicuous.

The second idea is that different parts of our scientific theories and practices have different historical dynamics. This means that they respond differently to the heterogeneity of historical factors. Thus, Lakatos proposed we talk of a hard core of assumptions that is more impermeable to revision than other tenets of a given theory. But the idea is more general than that. Even if we hold that all of science is socially constructed, that need not imply that the process of social construction is the same for every aspect and bit of scientific activity. The dynamics is much more complex than that. Heterogeneity is the rule. This is what Aliseda calls the “architecture” of scientific research.