I've recently been to a thesis writing course run by my university, and have been told that a thesis should include a references section (consisting of papers/sources you refer to directly) AND a bibliography (containing all the papers you have read during your PhD).

At this session there were only ~10 of us all from different departments, and a number of us questioned this, the reply being that sources you read but don't refer to directly may influence your thinking and therefore should be credited.

This is the only time I have ever come across this way of thinking, and I am unaware of a thesis that contains both. I'm just wondering if anyone here has done or heard this before?? As it seems impractical to me

I would check with your department specifically. Mine certainly did not have both. I only had a reference section for things I had consulted personally (and not simply read).

I think that it you put in everything that you had read over three years it would appear quite an incoherent mass (at least mine probably would!). A bit of discrimination of sources is a good thing!

I assume the "how to write a thesis" course was actually conducted by someone who did not have a PhD? Or was it a generic course for BSc, MA and PhD level?

Usually people only have references and Bibliography if they write an essay at Masters level. At PhD level, this is unlikely to be required as the purpose of this would be close to zero. You would have 300 references in your references and maybe another 1000 in your bibliography, accumulated over three years. That wouldn't make much sense.

Of all the PhD theses from my department I have seen; they only include a reference section for things that are specifically refered to in the text. When I mentioned this on the course I was told by the guy running it that this was "bad practice". The course was "How to write a PhD Thesis" run by the science and engineering department of the university's tranferable skills dept. On speaking with my supervisor and the department they say that, they require a reference section and if you choose to you can include a bibliography as well but that would be throwing yourself open to questions about any papers in the bibliography section.

But as the course was run by the university for their students I would have thought they should get things right

My university is crap in this, too. Often the courses are run by people who have no clue (and no PhD) but a lot of free time, have no value as they are to general and generic and sometimes (like in your case, apparently) they even give out wrong information.

I found courses run by my own faculty much more relevant and useful.

I would definitely only go for the reference section. I don't think a bibliography is meaningful for a PhD as Chrisr has said. The viva is a good point and you could indeed make things more difficult by including everything you have read. Realistically I never logged everything I read anyway from the start so I could not have done this.

Seems very bizarre to me.

I agree that Bibliography does not make sense for a PhD. If I saw an add in a women's magazine that gave me some idea for my PhD, then I have to reference it in Bibliography:)

I supposed it dependes on the area you are in. The majority of theses I read have a selected bibliography. Our Uni issued a guide that recommends to check with our dept.if we are expected to include everythting we have read (bibliography) or only sources we have used in the text (reference list). As far as I know we are expected to use the 'selected bibliography' and make a distinction only between primary (published/ unpublished) and secondary sources.

I think I will include both. Because the mixed advice I got from academics and reference guides confused me more, I decided to include a separate bibliography to cover my back really.

One of my supervisors suggests that I should include everything I read (cited or not) in the bibliography alphabetically. This way (because I'm using Harvard system) readers can cross examine the cited references easily but not be distracted with 2 separate lists. I have to get confirmation on that one from my other supervisors though.

hmm, that sounds even more confusing. So you would have only one list in total, but the majority of items listed are not part of your actual references but things you have read sometime? Quite frankly, I've never heard of something like this. Usually people have either one reference list or one reference list and an additional bibliography.

On the other hand, on the wider scale of things, these issues are rather minor and will probably not affect the outcome of the viva (I would hope).

I agree with you Otto. It sounds strange. I have seen some thesis with references only; and some with references and seperate bibliography. Not a one big list of everything. This supervisor has not done a PhD, but reads a lot and very a very bright person. So I am not going to dismiss that view totally but I see myself just adding a separate bibliography if necessary (that's how I enter info to EndNote, so it will be just a matter of sending it to Word.)