Here are some points/ideas I’ve found myself offering thesis-writing students over the years.

NOTE: This is an incomplete, highly subjective litany of questions, concerns, and suggestions. Reasonable people can disagree with any/all of the following. Consider, adopt, or discard as appropriate. Also, this may be updated periodically.

Any subjects you’d like to see addressed? Email me.

Preliminary Questions

Consider answering these questions as you get started:

• What is the genre of the text? Does the text exhibit traits of more than one genre?

  • Does the text deviate from the typical traits of its apparent genre(s)? If so, how?

• What is the significance of the text’s title?

• How have scholars typically understood/interpreted this text? What do they think makes it significant, in terms of its meaning, its cultural significance, and/or its contribution to literature?

  • How does your approach/analysis differ from other scholars’ readings?

• What literary, philosophical, cultural movement(s), school(s) of thought, etc., is the author typically associated with, if any?

  • How does your chosen text exemplify that association? How does it deviate from exemplifying that association?

• How recent is the theory you have chosen for your framework?

  • If the theory is not recent (ie, not from the past ~10 years), what are some responses to the theory, especially critical responses?
  • Is the theory typically used in the analysis of literature? If so, will your use of this theory follow or deviate from such uses?

• Are the theory and the text from different eras and/or parts of the world?

  • How might any difference in era/geography complicate the choice to use the theory for the analysis of the chosen text?

• Consider that the relationship between the theory and the text might go both ways.

  • Do ideas in the text challenge, build on, and/or differ from ideas presented in the theory?

Read a wide range of other theses — not just the ones you think are “good”

• You will be able to see more clearly what doesn’t work, and avoid doing that
• This is much easier than seeing what has worked and trying to emulate it
• This suggestion is based on the idea that we need to overcome “survivorship bias”; if we only consider the attributes of the “survivors,” we neglect to consider what led others to perish; we are then unable to see or consider those aspects in our own work, which might exist unseen alongside the positive attributes we’re emulating from the survivors.
• Reading reviews can also help you understand what doesn’t work, according to the reviewer

Identify what “they say” and build on it

• After you have read several texts on your topic, you can start to organize and categorize them, as a collective statement of “what they, the scholarly community, say”¹
• “To make an impact as a writer, you need to do more than make statements that are logical, well supported, and consistent. You must also find a way of entering a conversations with others’ views—with something ’they say.’”²
• This gives you something to expand on, or push against; this gives your work some meaning, relevance, weight, significance
• “[A thesis] must say something which one has not already explicitly thought or read but must be related in some positive way to what one has in the past thought or read. Within the context of what is known, it must propose modifications or elaborations.”³

Your paper is not just a collection of information

• Many papers are reading reports, (bad) descriptions of a few works of scholarship, followed by descriptions of the book in question, often using terms from the scholarship, or highlighting similarities between the scholarship and the primary work
  • Belcher warns students against “writing endless plot summary, rehearsing others’ theory, [and] stringing together tiny insights without any organizing principle”.⁴
• Consider that your primary aim is to teach your reader something exciting, interesting; inspire insight.
• “In a crucial sense, while the academic essay focuses on a subject, even more it’s an attempt to modify the audience’s perception of its previous relationship and response to that subject”.⁵
• “You may be attempting to discover ’new knowledge,’ yet you are also trying to persuade an audience not only that what you’ve discovered is important, but that it logically connects with, complements, resuscitates, clarifies—makes somehow vivid or vital or special—what the audience knew before.”⁶

Do not use a “discovery structure”

• Belcher writes: “Just because your precious insights took forever to arrive at, doesn’t mean you should force us through your process”.⁷
• Your paper should not follow the structure of a film, withholding the insights and epiphanies until the final act, to surprise your reader. Instead, give the game away right at the beginning, then use the rest of the paper to support the insight with facts and readings, showing that it is true, interesting, and important.

Write the whole time, not just in a flurry at the end

• You do not have to finish researching in order to begin writing
  • “Do not take endless notes and underline huge sections of books, and then feel overwhelmed because you have to go back through all of those notes and texts. Read and then write an actual paragraph, however loose, about what you have read”.⁸
  • I would add a caveat, that this advice seems to skip over the writing of literature notes; seems to suggest going straight to drafting. That seems hasty, and not advisable. Puts too much pressure on the paragraph, gives it too much weight.
  • “Writers who learn to leave holes in manuscripts to be filled in later master valuable skills in writing: they learn to proceed amid ambiguity and uncertainty”.⁹
  • “The best … researcher will be someone who learns to make a path through this immensity [of knowledge in the twenty-first century] without gettting overwhelmed”.¹⁰

Begin to structure your argument, in an ordered, bullet-pointed outline

• Use numbers to differentiate bullets, and nested bullets, eg. 1. ; 1.a ; 1.b ; 1.b.1 ; 2.
  • Doing so forces you to identify a thought, giving it a place; and it forces you (helps you) to distinguish between thoughts and lines of thought
  • Lines of thought can develop under a numbered bullet, by using sub-numbering
  • This allows you to begin, and continue developing, multiple lines of thought at the same time
  • Some of these lines of thought will not be developed very far
  • Some of these lines of thought will not make it into your final paper
  • But you won’t know how interesting a line of thought is until you’ve developed it a bit
• Developing a line of thought in bulleted outline form allows you to focus on thinking, not on formalities of writing
  • A bullet-point is always temporary, on its way to being final, never final on its own; this takes some of the pressure off; the bullet-pointed item can be a fully formed thought in fully formed prose; or it can be a fragment, a note, but it doesn’t have to be either one; if a fully formed thought in fully formed prose isn’t what comes to you, don’t stop your thinking to work on the writing. Just write the fragment and move on. You will be able to come back later and develop the line of thought, and the writing. It won’t get lost, because you’ve already given it a place in your outline. It has an address!
• For some of the first bullet points, just to get you started, write some descriptions and analysis of what “they say”
• Like the Dickinson poem: instead of going to heaven at last, I’m going all the while.

When I posted last week about tab-sets.el, I didn’t know that this month’s Emacs Carnival topic is “Obscure Packages.” Well, it’s hard to be more obscure than an unpublished package that I use for my own idiosyncratic ends. So, here’s another!

Introducing doc-tags.el: a simple, low-overhead Emacs package for tagging and retrieving files using a SQLite backend.

doc-tags is a straightforward package that lets you:

• Tag any file or directory on your computer, with one or more tags
• Quickly find and open files by tag, using “AND” (all tags) or “OR” (any tag) search
• Update, rename, or remove tags interactively
• Keep your tagged-doc database in a single regular file

You add only what matters to you: notes, readings, drafts, reference folders, syllabi, whatever. Tags are just Emacs strings — no renaming or relocating files, no global search, no magic.

Nota Bene

In general, I am skeptical about the long-term use value of tags. I find them flimsy. I much prefer organizing files in directories, and relying on searching to find what I’m looking for.

With that in mind, doc-tags is not — and is not meant to be — a replacement for file management writ large. I use devonthink for that. It is a huge directory structure of PDFs and lecture notes, organized within the devonthink program in all sort of useful ways, for long-term storage and findability, but otherwise — for all intents and purposes — a jumble of files in a maze of directories.

The major downside of devonthink, for me, is that its database structure is not accessible from inside Emacs. (Yet?) Sure, I can use a recursive search to plumb the depths of my devonthink database for what I think I need, but that’s a somewhat crude approach.

So, I wrote doc-tags as a middle-ground, ad-hoc solution, allowing me to assign tags to files in disparate locations across my file system, wherever they may be, even deep within my devonthink database.

Because of my deep skepticism of tags, I mostly use doc-tags when working on fleeting tasks, like preparing for a class or working on short-term writing projects. When the task is finished, I just delete the tags I no longer need and move on without remorse. No need to create a new directory, copy or symlink files. Just quick and dirty tagging, for the moment.

How to set it up

doc-tags.el isn’t up on M/ELPA, so install directly from GitHub:

(use-package doc-tags
  :ensure nil
  :vc (:url "https://github.com/localauthor/doc-tags")
  :custom
  (doc-tags-db-file "~/.emacs.d/var/doc-tags.sqlite") ;; pick your database location
  :config
  (doc-tags-connect) ;; opens/creates the database and initializes the schema
  )

You’ll also need triples (which manages the SQLite database and backups) and sqlite3 available to Emacs.

How does it work?

• To add a document: M-x doc-tags-add-doc (prompts for a file, then for one or more tags)
• To search by tag: M-x doc-tags-find-file (AND logic by default; use C-u for OR search)
• To add/remove tags: M-x doc-tags-add-tag / M-x doc-tags-remove-tag To edit/rename tags: M-x doc-tags-edit-tag
• To remove a document: M-x doc-tags-remove-doc

Most commands offer smart completion (minibuffer with tag annotations, document previews); if you use embark, doc-tags offers a keymap for quick actions on tags and docs.