When Kuhn and others questioned realism in science, the “tool” of reference was used to argue that the world is not well lost.
The following passage expresses what was at stake after Thomas Kuhn’s own “revolution”:
“[W]e need an alternative account of meaning which allows that people holding competing or successive theories may still be talking about the same thing.”
Partly as a reaction to what many took to be “Kuhnian relativism”, some philosophers wanted to find a means of fighting back. They did so by focusing on reference.
Meaning always (or often) goes alongside reference in this tradition. But “what then is the constant in meaning?” According to the historian and philosopher of science Ian Hacking, Hilary Putnam “makes everything turn on reference and extension”.
The most obvious thing to say here is that these philosophers were attempting to securely tie names, words and terms to the world.
Fregean Sense
Long before Kuhn and Putnam, Gottlob Frege “thought that a word has a standard sense, which is what makes a scientific tradition possible”. What is that sense? —
“The sense is what is shared by all communicators, and may be passed down from generation to generation of students.”
It will be shown that Hilary Putnam’s “stereotype” (to be discussed in a moment) shares a family resemblance to Frege’s sense. (They’re still very different notions.) Still, what is important here is that Frege
“taught us that an expression should have a definite fixed sense, which we apprehend, and which enables us to pick out the reference”.
In other words, the route a word/term takes to a thing goes via this fixed sense.
Hacking’s Own Position on Reference
Sometimes it’s difficult to decipher if Hacking is putting his own view or that of someone else (e.g., Putnam). In this case, he puts the case for the permanence of reference in the following way:
“If we do have a genuine natural kind term, the reference of the term will remain the same, even though stereotypical opinions of the kind may change. Thus the fundamental principle of identity for a term shifts from Fregeian sense to Putnamian reference.”
We’re told that the reference (miraculously?) remains the same even though what are deemed to be its stereotypical properties change. The claim here isn’t that the reference can be “found” without relying on its stereotypical properties, only that such properties change. The obvious question now is:
If the stereotypical properties are always changing, how do we know we have the same reference?
The Putnamian position is that we get at the reference via his version of “Fregean sense”. The Fregean sense “may change”, while the reference, qua reference, doesn’t. However, does that really answer the question (in bold) above?
To be clear. Hacking accepts the importance of reference for science. However, he tells his readers that “[a]nybody who thinks that reference works by a causal and historical connection to the thing named ought to reflect on the following example”. Hacking then gives a complex historical example in which the notion of “causal and historical connection to the thing named” doesn’t work, but Hacking’s own entity realism does.
So this isn’t an attack on reference. It’s an attack on a particular account of reference.
Hacking on Acid
Hacking provides a concrete example of possible incommensurability as it relates to the word ‘acid’:
“The incommensurabilist would say that we mean something different by the word ‘acid’ than did Lavoisier or Dalton around 1800. Our theories about acids have changed substantially [ ].”
Of course, this is very difficult for the layperson to grasp because he’d need to have both historical knowledge and a knowledge of chemistry. It would work very differently if, say, Hacking gave as an example the word ‘cat’ or ‘brick’. Yet even the layperson would probably accept the clause “our theories about acids have changed substantially”.
In any case, Hacking continues:
“[ ] Putnam says, we are still talking about the same kind of stuff as those pioneers of the new chemistry.”
Intuitively, and simply, this seems feasible. It could be a case of people saying different things about the same stuff. Even contemporary persons could be saying different things about the same stuff.
This position posits a kind of strong distinction between the stuff itself and what we say about it. Can that distinction be maintained? Wouldn’t it assume that firstly we have uncontaminated stuff, and, later, our sayings about it? What if that stuff is always contaminated with sayings or theories? That may be the case. Yet the same stuff would still underpin all these different sayings and theories.
Reference and extension are means to hold on to that same stuff.
Putnam’s Stereotypes
According to Hacking, Putnam justified his position in terms of what he called a “stereotype”:
“Certainly there is an important cluster of properties and in the professional stereotype for acids: acids are substances that in a water solution taste sour, and change the colour of indicators such as litmus paper. They react with many metals to form hydrogen, and react with bases to form salts.”
That’s a decent number of properties we can use to identify acids. Is this a roundabout way of saying that everyone in the know will experience a sour taste, experience a litmus paper changing to a certain colour, etc. regardless of whether they lived in 1800 or 2020? This would mean that any disagreements about acids must occur elsewhere — beyond these specific properties. It seems that the differences in theory account for these differences.
Hacking mentions theory in the following:
“Lavoisier and Dalton would agree completely with this stereotype. Lavoisier happened to have a false theory about such substances, for he thought every acid had oxygen in it.”
So we have different theorists all experiencing a water solution tasting sour, indicators such as litmus paper changing colour, stuff reacting with many metals to form hydrogen, and the stuff reacting with bases to form salts. Over and above all that, we have theory too.
(Just to get this point across, let’s take Putnam’s other example of the word ‘water’. Putnam “gives us colourless, transparent, tasteless, thirst-quenching, etc [as] part of the stereotype for ‘water’”.)
One question readers may ask here is whether or not all these properties (or stereotypes) really are theory-free. After all, we’re taking about hydrogen, oxygen, litmus paper, salt, etc. here. Perhaps that doesn’t matter in this specific context.
Hacking concludes with a entity-realist statement: “But there is no doubt that Lavoisier and Davy were talking about the same stuff.”
Davy is mentioned here because in 1810 he showed that Lavoisier’s theory “was a mistake”. That is because “muriatic acid is just HC1, what we now call hydrochloric acid”. Here again we’re shown that theory rides piggyback on these stereotypical properties of the stuff. Indeed, the new term “hydrochloric acid” does so too.
These general points are made again by Hacking we he states that
“the philosopher of naming must ask if Lavoisier meant Brønsted-Lowry acids or Lewis acids when he spoke of acids.”
Here’s the crunch: “Obviously he meant neither.”
That means that there is a distinction to be made, again, between stereotypical properties and theories. Can we now say that Brønsted-Lowry acids and Lewis acids are theory-laden, but stereotypical properties aren’t? (The philosopher Russell Hanson, way back in the 1960s, would have certainly said, ‘No, we can’t.”)
Bas van Fraaseen’s Electron
A much-better known case of naming is the electron. Hacking writes:
“[Bas van Fraassen] does tease the realist, who is confident that there are electrons: ‘Whose electron did Millikan observe; Lorenz’s, Rutherford’s, Bohr’s or Schrodinger’s?’”
On our theme of theories, isn’t the Dutch-American philosopher Bas van Fraassen referring to the different theories of the electron, not to the electron itself? After all, don’t we have the phrase “different theories of the electron”, not “different theories of the banana”. That said, what the hell is this theory-free electron? There’s nothing more theory-laden than the electron, surely. Tell me, what is the electron without theory? It was born with theories attached to it. It has never been free of theory. The realist, or Putnam, may concede all this. But that still leaves stereotypical properties which run free of theory… Or do they?
Now take the (entity) realist’s “obvious reply”, at least as expressed by Hacking:
“Millikan measured the charge of the electron. Lorentz, Rutherford, Bohr, Schrodinger and Millikan were all talking about electrons. They had different theories about electrons. Different stereotypes of electrons have been in vogue but it is the reference that fixes the sameness of what we are talking about.”
It’s not just that theories about any given x change, so too do the stereotypical properties. Hacking is saying that we have something, “the reference” (or the electron), which is free of not only theory, but also any fixed set of stereotypical properties. Isn’t it this position that really irks the anti-realist or instrumentalist? What can we say about the reference without any stereotypical properties or theory? Perhaps, then, Hacking isn’t arguing that we do have the reference (or electron) without any stereotypical properties. It’s just that such properties can change. That is, they can be “in vogue” this week, and out of vogue next week.
We can now take a slightly different approach to the importance of reference — one that doesn’t actually focus on scientific terms.
Donnellan’s Causal Theory of Reference
The philosopher Keith Donnellan attempted to save philosophical realism via reference. In other words, if we can have a causal theory of reference which excludes names like ‘Santa’, then we can distinguish real propositions like “Lady Gaga will come tonight” from “Zeus is angry with us”.
Donnellan argued that proper names shouldn’t rely on being tied to “descriptive content”. Metaphorically, he believed that this would give individual minds too much power, and direct causal reference less power.
Associating descriptive content with proper names has been deemed to be idealist by some philosophers too. (“Subjective” or “mentalist” is a better term in this context.) A causal connection to Lady Gaga (or the ‘Lady Gaga’ relation to Lady Gaga) is concrete. The descriptive content “The rubbish [or great] singer in Hollywood” is subjective, or at least dependent on an individual mind (or on individual minds in the plural). Similarly, the definite description that readers personally tie to the name ‘Santa’ may be very different to the ones other people tie to that name.
So what would secure us identity of reference here when it comes to the names ‘Santa’ and ‘Lady Gaga’?
Donnellan’s causal theory of reference (which excludes Santa) is the realist’s way of avoiding not just idealism (or, better, subjectivity), but also relativism.
Conventions, conceptual schemes, identifying descriptions, etc. were deemed to be subjective by Donnellan. Causal processes, on the other hand, are deemed to impose themselves on us. In other words, they tie us firmly to the world.
