How much is too much when defining and explaining terms in depth in a dissertation?

Question

When writing a dissertation for a PhD in computer science, where the research is interdisciplinary, is it OK to define and include in depth explanation for an extraordinarily large amount of terminology in the preliminaries and background?

Currently, in my preliminaries and background section, which follows the introduction, I have 50 definitions but will have more, many of which are trivial or common knowledge to some readers. Each definition comprises a substantially long paragraph. My reasoning and upfront explanation is that I hope to make the dissertation more accessible, friendly to non-English first language readers, and that the interdisciplinary content requires more definitions and explanations than usual to make the dissertation understandable to those in my field.

Answer 1

Please make your dissertation welcoming to those not in your specialty and to those who do not share your language fully.

I often read dissertations when I am trying to get to know a new subject. I know many people read my dissertation, as it picked up lots of citations over the past few decades. I have no idea why people say only a few people will read your dissertation. That seems insulting. I have read a few undergraduate dissertations to find some formulas spelled out. More commonly, I read masters dissertations, and many more PhD dissertations, mostly those in fields that were not where I got my graduate training.

The exception is if you are in a harsh field where being nice to your readers is seen as a sign of weakness. Even then, you might be brave and buck the trend. Most importantly, do listen to your advisor. Don't spend forever writing this. Often the longer version takes no more time as you need not decide what to cut out, but at some point you need to pick a style and power through. Your committee can tell you to add detail or take out detail.

My advisor taught me that a good mathematician writes a "full-proof" version of every paper that nails down every detail, and then edits that down to make the published version. He wanted my dissertation to be the full-proof version of a few papers. Later, one is to only publish the short and efficient versions, but had the full version on a shelf to pull down if someone questioned you on a point.

If your dissertation is structured correctly, your committee can skip over sections that give background that they know well. Your advisor will get stuck reading it all. One does get paid for this, however.

Answer 2

Generally in order to be reader-friendly, it is more advisable to explain too much than too little. If you explain not enough, some parts of your thesis may be incomprehensible, which devalues them. If you explain too much, people still can understand your stuff, even if they may lose a bit of time. For sure "explaining too much" is not a reason to fail your PhD (unless there is something like a word count limit and an excessive account of basics stops you from explaining you own work properly).

The only problem I see is that there may be the odd expert reader who is annoyed about having to read so much they already know (they may even believe everybody who reads the thesis will know this-or-that, whether that's true or not), and this may give them a certain negative feeling when going on reading. This however should not normally be a problem; the quality of your original work is what it is, and explaining too much doesn't affect the readers' ability to see this quality. This is what you should be assessed on, so I believe it can't hurt, even though personally I'd for sure ask your supervisor and maybe others for feedback about that stuff; even if it won't make your thesis fall, if being nice to your main readers means explaining a bit less, I'd recommend to be nice.