I do not think that we disagree. I over-estimated how far you wanted to take your division of labour. And with yesterday’s announcement that funding for universities is going to be cut, efficiency is going to move sharply up the list of priorities.

Richard in not-so-sunny London

Your proposed division of labour might lead to greater efficiency. Division of labour often does. But I worry about taking the division right back into the training process. Can theoreticians be relied upon not to go way beyond what the evidence will support, if they do not know how the evidence is gathered? Can the gatherers of evidence be relied upon to notice the right things, if they do not know how the theoreticians will use their work? And who will have the inter-disciplinary skills to challenge new or established theories by reference to the evidence, or to propose new ways of gathering evidence in order to facilitate the advance of theory?

Again, I think it is interesting to draw comparisons with other disciplines. In at least some parts of the natural sciences, there are experimenters, and there are theoreticians who must not be let near a laboratory because they are bound to break something. There, the division of labour works well. But that might be at least partly because the claims of theory that must be tested by experiment, and the experimental results that call for theoretical interpretation, are precisely defined in mathematical terms. I do not think that the humanities have this advantage.

Absolultely! It is also about creating the tools to ‘excavate’ the Web and collect useful and relevant resources.

See also this website that aggregates live online content on museum archaeology and archaeological heritage:

http://mashuparchaeology.humanities.manchester.ac.uk/index.php

But technology should be used cautiously. A few weeks ago I was going through a book looking for somthing, and I caught myself wondering “Where on earth is that ‘search’ button?”!

But there is still a case for making students perform routinely automated tasks manually, and routinely computer-aided tasks the hard way, at least a few times. They need to learn what our helpful computers do, and on the basis of what information (the features of an artefact, or the keywords in a literature search). There are at least two reasons for this. The first is that they will come to understand the limitations of what computers do for us, and in particular the kinds of things that they might miss. The second is that a student who knows what a computer does is in a good position to spot ways to go beyond the results that the computer gives, and indeed to write the next generation of programs and databases, or at least specifications for them.

The need for students to perform tasks the hard way is, I think, most obvious in mathematics. If, for example, a student knew how to give a computer functions to integrate, but had never worked out some basic integrals from first principles, could he or she be said to understand the computer’s answers? Turning to numismatics, if a student gave a computer an image of a coin and got the answer “Roman, minted in Gaul about 200 AD”, but did not understand how the computer reached that conclusion because he or she had never dated a coin the hard way, he or she might understand the answer so far as it went, but could not be said to have that answer rooted in a general understanding of numismatics. He or she would therefore only have a very limited understanding.

There are interesting questions here about degrees of understanding, and the extent to which knowledge of a wider context and of the methods by which answers were reached can be said to be necessary for understanding. I shall ask my philosophical computer to answer these questions for me, and shall meanwhile raise my hat to Plato for his suggestion (although not his conclusion) that knowledge requires a rational account, in addition to true belief (Theaetetus 201-202).