So we ended yesterday on the note that there are no laws in the social sciences, according to Roberts, because hedged laws aren’t really laws. Thankfully, Harold Kincaid disagrees. He says there ARE laws in the social science, because social sciences study societies which are made up of human beings which are physical entities. Physical entities are governed by laws that physics describes, therefore… there must be laws that govern societies (and not just ones that we make up in legislature).
Kincaid also asks, “what is a law?” He answers it first by saying that they’re exceptionless generalisations. Uh-oh. That doesn’t sound good for social science… Kincaid says yeah, there has been a lot of debate, but laws are actually statements that identify a causal factor. This does seem to set the bar for lawhood pretty low, if anything that identifies a causal force has met the sufficiency criterion for being a law.
To the first objection, which says that laws are exceptionless generalisations, Kincaid points out that no regularity is exceptionless. Pretty fair. The inverse square law identifies a causal force (gravity), but ignores all other forces, so it clearly does have exceptions. We call it a law, but we seem to forget that it’s kind of a hedged law.
Second, there is the argument that laws aren’t just accidental generalisations, and some accidental generalisations may pick out causal factors. Additionally, laws are supposed to tell us what would happen if things are different (i.e., support counterfactuals). Statements that just identify causal factors don’t do this. Laws also allow us to make predictions, and supposedly causal factor’s don’t. In fact, causal factors don’t seem to support counterfactuals, either. Right?
Wrong! There isn’t really a division between laws and other types of causal claims, Kincaid says. Every causal statement given has a degree of necessity, and they also always have some counterfactual implications. What differs is their scope or breadth; generalisations in the social sciences that involve picking out causal factors may have smaller scopes than other laws, but they still have some modal breadth. Causal claims can also support some predictions, so this objection fails.
Another objection is that laws describe very fundamental causal claims – ones that aren’t the result of much deeper forces. This is kind of a bad objection, though, because even within the physical sciences there are derivative laws that still count as laws because they explain and predict stuff.
Yet another objection is that laws don’t need to have causal claims, to which Kincaid says fine. Not all laws have causal statements, but causal claims are a type of law nevertheless. Objections that try to separate causal statements from laws will fail.
In the end Kincaid says that laws cite causal factors that help us explain and predict – but not all laws have causal factors, and that’s fine.
“The key question thus is whether the social sciences provide causal claims that provide relatively extensive explanations and predictions.”
If they can do that, then the social sciences have laws and are legit sciences. How are causal claims established? By “tak[ing] background knowledge of causes, observ[ing] various changes in factors of interest, and infer[ring] what causes what.” If we can show similar effects in repetitions, “[s]uch knowledge is strengthened and deepened.”
Do social sciences meet this requirement? Social sciences often employ experiments that are qualitative in nature, and they also make background assumptions that are patently false, so it seems like they don’t meet this requirement. Kincaid knocks this down by saying that we don’t necessarily have to measure how much one factor changes when we alter an independent variable. We can attribute results to the manipulation of some factor when we hold everything else constant, and we also clearly have many sciences that don’t rely on strict scientific experimentation (e.g., astronomy, paeleontology, geology, evolutionary biology).
What about the assumptions bit? We know these assumptions are either idealisations or abstractions, but if they are false why does an experiment based on them make any difference? Well, for one thing, the natural sciences often freely employ false idealisations and abstractions. The ideal gas law is so-named because it works under the assumption that a gas behaves in an ideal fashion, EVEN THOUGH we know that gases do not always behave ideally (i.e., we have certain circumstances under which we know gases do not behave ideally). We still know that the causes mentioned in the theory are what are responsible for data, and not the falsity of the assumptions. Nonexperimental natural sciences show this, and social sciences do as well.
Kincaid uses the law of supply and demand to demonstrate this, and points out that the law has a causal factor identified (price). So it’s pretty law-like. In fact it’s a law. It’s hedged, but it’s still a law. It does not say that price is the ONLY causal factor, because there are certainly other events that are going on that will affect demand or supply, but the crucial thing is that it identified a causal factor. Multiple lines of observational evidence seems to strengthen the law, also, so there we go. Economics is a social science and it’s got a law… so there are laws in the social sciences.
Any objections relating to the ceteris paribus clause associated with the law of supply and demand seem to have no water given previous points mentioned about this. Laws in the physical and natural sciences also have ceteris paribus clauses, but we still consider them laws. Most sciences deal with complex phenomena, yet still produce causal knowledge that help explain and reliably predict stuff.
There’s an objection that if there are laws in the social sciences then we have reduced human beings to automatons devoid of free will. Again, this is not true; social sciences may have laws, but they do not solely use laws to explain phenomena. Only the libertarian’s notion of free will is problematic in this case, because the compatibilist does not claim that free will and causal influences can’t coexist. (Only the libertarian says that choice is 100% uncaused by anything else aside from the person making the choice.)
Kincaid finds the libertarian notion of freedom troublesome, because it would seem to imply that if we are the only cause of our actions that we must have created ourselves. Obviously this isn’t what happens… Self-creation is not possible for humans. You can make something of yourself, but you do not literally make yourself (especially not from scratch!).
Kincaid further says that the objection fails because it “presupposes that the social sciences are only about individual behaviour”, which they are not. The social sciences may also be about collective behaviour, because in reality no man is really an island and we act differently in groups than we do alone (oftentimes).
In the end (I left out one objection that Kincaid also takes down, but I’m tired of writing and this is so long already), I think Kincaid’s argument is pretty convincing. Of course, it’s a little difficult to say that he or Roberts are actually debating the same thing… After all, they are working with different definitions of what it is for something to be a ‘law’. There is also a question of whether one needs advance knowledge of what counts as interference if one is to have a law, but Kincaid says you don’t (but even if you did you could supposedly identify interference for the law of supply and demand).
While reading Kincaid’s paper I did keep wondering if he was relying too much on the law of supply and demand as an example. Is he pushing the analogy a bit too far? Even if gravity is a nicer law than the law of S+D, it doesn’t defeat the idea that there are laws where there isn’t vector addition, so it’s not a huge problem that he pushes the analogy in his argument.
Probably the biggest problem for Kincaid is whether he sets the bar too low for laws and what it means to be a ‘law’. Where exactly is the line between laws and non-laws anyway? We can have a law that has causal factors and a broad scope, but when we narrow the scope down but it maintains its causal factors does it become a non-law? What about when we eliminate the causal factors?
Questions, questions. Philosophy is so complex…
Scientific Student 