Consider the following example:
Case 1: ‘Tremulous Hand’
‘Tremulous Hand’ is used to refer to the otherwise unidentified author of around 50,000 Thirteenth Century glosses in manuscripts. Palaeographical analysis provides strong evidence that these glosses are the work of a single person with distinctive (tremulous and left-leaning) handwriting. All that is known about Tremulous Hand is what can be deduced from the glosses themselves.
‘Tremulous Hand’ is a descriptive name: a name associated with a stipulation of form ‘Let NN refer to the F’, and associated with a characteristic way of justifying beliefs you would use the name to express (move to <NN is G> given justification for believing <The F is G>). Most people who have written on descriptive names have agreed that the aboutness-fixing story for the thoughts we use these names to express is ‘satisfactional’: the thoughts are about the satisfier of the associated description. The debate has then been about whether we should count these thoughts as genuinely ‘singular’.
The Fixing Reference framework generates a different account of how aboutness-fixing works in a ‘Tremulous Hand’-type case. In this framework, a body of beliefs associated with a given means of justification is about an object iff it is the object on which the means of justification converges – the object whose properties you’ll be unlucky to get wrong if you justify beliefs in this way. A body of beliefs expressed (or regarded by the subject as expressible) using a descriptive name is associated with a specific means of justification: move to <NN is G> given good enough evidence for believing <The F is G>. So, in this framework, the body of beliefs is about o iff o is the unique thing whose properties you’ll be unlucky to get wrong if you use this method of justification to form <NN is G> beliefs.
One immediate consequence here is that whether the beliefs expressed using a descriptive name are about an object comes apart from whether the object satisfies the associated description. Here’s another example to illustrate this possibility:
Case 2: ‘Geraint the Blue Bard’
‘Geraint the Blue Bard’ was used for over a hundred years as a name for the otherwise unidentified author of a series of songs in medieval Welsh, dealing with medieval themes, and employing medieval metres. Efforts to find out more about Geraint’s life, taking off from cues in the texts, supposed that he flourished in the ninth century, and was either an apothecary, a minor aristocrat, or a priest. Rival factions collected large bodies of evidence to support each of these hypotheses. But in 1956 the ‘Blue Bard’ songs were shown to be the work of notorious nineteenth century forger Edward Williams.
(This is a real example – pronunciation of ‘Geraint’ here: https://downloads.bbc.co.uk/rmhttp/wales/livinginwales/geraint.mp3)
I take it that there is an intuitive verdict about this case: there in fact was no Geriant; the beliefs the scholars working before the discovery of the forgery used ‘Geraint’ to express were about nobody. In the Fixing Reference framework, the aboutness-failure in this case traces to failure in cognitive focus: given how the scholars justified their ‘Geriant’ beliefs, it would have been a matter of spectacular chance if they’d ended up with beliefs that matched Edward Williams’s properties, so the beliefs the scholars used ‘Geraint’ to express were not about Williams even though he was the satisfier of the description associated with name.
The way I’ve put things so far in this post, and the way I treat this material in the book, stresses the similarities between the aboutness-fixing stories that underlie our uses of perceptual demonstratives, on the one hand, and descriptive names, on the other: in each case, aboutness-fixing is a matter of securing cognitive focus. But it’s also important to recognize that there are different kinds of focus involved in each kind of case.
To see the difference in intuitive terms, think about the contrast between an optical telescope and a radio-telescope. For an optical telescope, focus is a relation to an object which contributes to shaping the information signal that the telescope delivers: it is a relation that secures the result that this signal will match the object unless some unlucky spoiler intervenes. For a radio telescope, in contrast, focus is a relation to an object that is generated by post-signal processing. The radio telescope generates a noisy buzzy information feed which is then crunched (this is a technical term) using algorithms. If the astronomers have done their job properly, the crunching yields a report which matches what some object in the sky is like except in respects in which the object is behaving in very unusual ways. In this case, the astronomers have attained focus on the object. But the focus is generated by post-signal processing, rather than itself being part of what generates the signal.
Though I’ve only just begun to explore this idea, my current view is that the cognitive focus that underpins our use of perceptual demonstratives is analogous to focus for an optical telescope, while the cognitive focus that underpins our use of descriptive names is analogous to focus for a radio telescope. Though both perceptual demonstrative thought and descriptively mediated singular thought are genuinely singular, this difference in the kind of cognitive focus involved reflects a deep difference in the roles these kinds of thought play in our cognitive economies.
(The last two paragraphs here have been prompted by a series of excellent comments on this part of my view from Karen Lewis, Calvin Normore, and James Shaw).
In the Geraint example, this is the standard fiction setup: G. is a concept in the mind of Edward Williams that scholars beliefs are about. As to telescopes, what about Lowell and canals on Mars? Many visual perceptions can be shown to include estimates of their vagueness/uncertainty (at the neural level), which need to be ignored before we accept a false belief.