Safe, Safe Baby

So, testimony. My last proper (read: contentive or content-present) post was all about whether testimony could be considered properly generative (on its own account). My thesis was just a tad hazy, as can be discerned by the questions raised by two of my earnest commentators. As this particular angle on testimony isn’t likely to come up in the paper I am writing I’m going to justify this blog’s existence by bleating on a little longer. As is my manner (and as is my right) I’ll do it in a fairly circuitous manner.

I want to talk unsafe transmission.

Yes, verily, ‘All-Embracing…’ is all about the sex, baby.

[Normally I’d put a (more) tag here, but I’ve currently given up on hiding the length of my posts, mostly because my list of ‘Pages’ is now so long that I need full-length posts to hide the fact that my sidebar is now unwieldy[1].]

Traditionally it has been held that testimony is only successful when someone who holds a justified belief is able to transfer that belief to you with the same justification. Thus A tells me that ‘Julius Caesar was assassinated on the Ides of March’ and as I know that A is an historian and thus gets her information from the right sources I believe that ‘Julius Caesar was assassinated on the Ides of March’ as well, taking onboard A’s justification as my own (in a sense). We don’t usually believe that testifiers who have unjustified beliefs can transmit those unjustified beliefs as testimony (although that gets a little murky when we consider that some of the things we used to believe in, such as astrology, were quite complicated beliefs, transmitted by learned individuals, treated as knowledge).


Yet there is a case for claiming that unjustified testifiers can cause justified belief in hearers. This is unsafe testimonial transmission. I read a paper on this by Sanford Goldberg and it stirred the inner juices of my brain cavity. I think a lot of my suspicion about the generative nature of testimony came out of this one paper, so perhaps, to save face, reduce confusion and create obfuscation, I should reveal all.

Goldberg uses the idea of local invariance to account for unsafe testimonial transmission. To illustrate this he provides a case very much like this one:

Frank has a peculiar habit of buying milk every day and then, no matter the circumstances, pours it down the drain the following morning before putting the empty carton back in the fridge before he sets himself to work in the very same kitchen. Staying with him are Mary and her child Sonny. One morning Mary goes to get some orange juice and sees the milk carton. By pure accident Frank has forgotten to perform his usual disposal and thus there is, actually, milk in the fridge. Mary tells this to Sonny when he asks if there is milk in the fridge. Goldberg claims that this is a case of unsafe transmission because, although Mary has an unjustified belief that there is milk in the fridge, Sonny gains a justified belief.

Why? Well, because:

Had there been no milk in the fridge, this would have been because Frank dumped it (and put the empty milk carton back in the fridge.) As noted above, in such a situation Mary would still have testified as she did; but Frank (who is a fixture in the kitchen, and so who is in the kitchen in most or all of the nearest possible worlds) would have immediately spoken up against the testimony, informing his uninitiated guests of his strange practice. In that case Sonny would not have consumed Mary’s testimony and so would have refrained from forming the testimonial belief that there was milk in the fridge. This establishes that Sonny’s testimonial belief is sensitive. Now, had Sonny formed the testimonial belief that there is milk in the fridge, this would have been a case in which Frank did not speak up against that testimony; but, given Frank’s scrupulousness, the only cases in which he would not speak up against that testimony (given that he was in the kitchen, as always) would be those cases, like the actual one, in which (upon hearing the testimony) he came to acknowledge that he failed to dump the milk from the previous evening. In all such cases, there would be milk in the fridge. In sum, had Sonny formed the testimonial belief that there is milk in the fridge, there would have been milk in the fridge: Sonny’s testimonial belief is reliable. Note, too, that any nearby world in which (a) Frank disposed of the milk and returned the empty carton to the fridge, yet (b) Mary – or someone else, for that matter – testified (on the basis of seeing the milk carton in the fridge) that there was milk in the fridge, will be a world in which Frank speaks up against that testimony, prompting Sonny to refrain from consuming that testimony. Sonny’s belief is safe.
-Goldberg, Sanford, ‘Testimonial knowledge through unsafe testimony,’ Analysis 64:4, 2005. p. 303-4

Frank is an example of local invariance; he is a fixed condition of the world Mary and Sonny inhabit. Had there been no milk in the fridge he would have spoken up, thus defeating Mary’s testimony. The fact that he did not speak up means that Mary’s testimony, although for her unjustified, was justified for Sonny because of Frank.

It’s a curious little case study and it introduces a whole new further issue in the relationship between a speaker and a hearer. Although it took a while to become formalised in the literature, we have worked out that the relationship between trustworthiness of sources and the truth of the proposition they assert is not necessary for testimony to be good. Untrustworthy people can assert the truth and trustworthy people can have momentary lapses of reason. Goldberg’s thesis suggests that trustworthiness of sources can be considered irrelevant in certain cases; what really is important is that the world functions properly (or reliably) rather than the agents within it (which makes sense, because Goldberg’s thesis is firmly centred on a reliabilist account of epistemology, where the proper function of processes is all important).

So, back to generation. A really ideal example of generative testimony (in re that last, contentaive post) would have the new knowledge come out of unsafely transmitted beliefs. I keep thinking about Math, mostly because there are lots of examples of what we would call mathematical knowledge generated from the contradictions of previously well-held views. I am also suddenly contemplating Galileo and his experiments to do with mass; before he performed the ‘Ball drop from Pisa’ he had already worked out that the Aristotelian model of mass-cum-weight was contradictory (heavier objects linked to lighter objects of the same mass should drag the lighter objects down in free fall but, importantly, lighter objects linked to heavier objects of the same mass should drag the heavier objects down in free fall; which was it? Well, it was neither…).

So, yes, more thoughts on the subject. Because thinking is good, unless you only think you are thinking, which is just all too common.

1. Unfortunately the theme I am currently using for WordPress is a just a tad incompatible with the version of WordPress I am using and so it only is held together with paperclips and bubblegum. Thus the nice plugins that would make the categories and the page links fold-away magically don’t work. I’ve yet to find that perfect theme replacement, so kludges it is.

The Generation Game

There is a common claim in the epistemology of testimony, which is that testimony requires witnessing and so isn’t generative of beliefs in the same way that memory is. Memories can combine, it is claimed, to generate new beliefs (you remember that your parents told you the cat went away to the farm and you remember another instance (when you were older) where your parents playfully suggest to some other parents that they should tell their kids that the cat wasn’t put down but was sent away; lo and behold you generate a new belief that your parents cannot be trusted) but that testimony simply transmits existing beliefs between the speaker and the hearer (thus any new beliefs are generated not by testimony but with testimonial support acting along with beliefs you acquired by other means).

I’m not sure I find this (above, crude) account all that convincing. It is true that for a lot of beliefs that are transferred via testimony there did have to be an original witness. For example, a lot of us know about the structure of the atom yet, I would wager, most of us have not performed the necessary experiments to know ‘first hand’ (if you will) that structure. We have, however, been told about it, by reputable sources. Yet I’m also fairly sure that a lot of beliefs have been generated via testimony as well. I’m thinking here of emergent beliefs; beliefs that have come out of inferences from other beliefs. Take the atom example. A lot of the information we have about atoms originated from experimental data; someone ‘saw’ something and then told others. But surely some of the subsequent beliefs about atoms have been generated by people who haven’t performed the experiments but rather have taken these reliable bits of data and inferred further beliefs (which then may well have been subject to testimony). I was going to use an example from mathematics but as it is really quite hard to tell a story about witnessing the truth of any given mathematical proposition I’m not entirely sure this is a good avenue to explore. In any case, surely there are cases of people using only testimonial beliefs, to which they have no other support, to generate new beliefs (which will then count as knowledge) which can then be transmitted testimonially?

The notion of testimony generating belief is contentious as it stands. Reductionists, those who hold that testimony is a second-class citizen compared to perception, memory and the other two classically recognised senses, claim that testimony can’t generate beliefs because experiences generate beliefs and testimony is simply belief-transfer (criude but basically true). If I am right, however, then the reductionist needs to distinguish between primary and secondary testimony. Primary testimony is non-generative; the proposition involved was generated by one of the four senses. Secondary testimony is generative, however, in that the component beliefs of secondary testimony (pieces of primary testimony) infer the belief that is the secondary testimony.

Non-reductionists (oft called anti-reductionists (you can guess by who)) accept that testimony can be generative (via different accounts, usually), in which case you have primary testimony which is generative by whatever mechanism produces testimony and secondary testimony, which is the inferences licensed from primary testimony (so secondary testimony is just a special class of primary testimony).

So, the question is, can I come up with some good examples of secondary testimony? Or, because this is a blog, can you? My problem is that I suspect that most of the obvious candidates for my theory are probably going to be examples of primary testimony. What I really need is an example of an idea where the belief being transmitted really isn’t based upon someone inferring something as an hypothesis, testing it and then transmitting their belief. It needs to be an example of taking primary testimony and, from that alone, coming to knowledge of something inferred from it that doesn’t take its justification (for its truthiness) from the primary testimony that produced it.

Ideas, anyone?