Why academic prose is hard to read, the paragraph habit that fixes most of it, and how to write when you do not feel ready.
Most difficult academic prose is not difficult because the ideas are complex. It is difficult because the writer was unsure and hedged, or wanted to sound scholarly and reached for abstraction. Clear writing in a thesis is largely a matter of a few habits: one idea per paragraph stated at the top, sentences that name who does what, and enough signposting that a reader always knows where they are in the argument.
Three habits produce most of it, and all three are learned by imitation from published work that also has them.
Hiding the actor. “It was determined that a relationship existed” hides who determined it. Passive constructions have legitimate uses — when the actor genuinely does not matter, or when your discipline requires them in methods sections — but used throughout they make prose vague and oddly evasive. Where your field permits it, saying “I selected” or “the study found” is clearer and increasingly accepted.
Nouns doing verbs’ work. “The implementation of the intervention resulted in an improvement in engagement” is three actions turned into nouns. “Introducing the intervention improved engagement” says the same thing and can actually be read. This single change fixes more thesis prose than any other.
Hedging beyond the evidence. “It could possibly be suggested that there may be some indication of a potential relationship” is four hedges around one claim. Hedging is necessary and honest — you are frequently uncertain and should say so — but one hedge is enough. Pick the level of confidence your evidence supports and state it once.
None of this is about dumbing down. Technical vocabulary is precision, and using the right term is clearer than explaining around it. The problem is abstraction where a concrete word would do, and complexity that came from the writing rather than the idea.
If you change one thing, change this: every paragraph makes one point, stated in its first sentence.
Most difficult thesis paragraphs contain three or four points, arrive at their main claim in the middle or at the end, and leave the reader assembling the argument themselves. Supervisors describe the result as “unclear” or “hard to follow” without always identifying the cause.
The structure that works: state the point, provide the evidence, explain what the evidence shows, and connect to what follows.
Two diagnostics you can run on your own draft:
Read only the first sentence of every paragraph in a chapter. Together they should form a summary of the argument. If they do not, either the paragraphs are not leading with their points, or the argument does not have a sequence.
Try writing a one-line summary of each paragraph in the margin. Any paragraph needing two lines is doing two jobs and should be split. Any paragraph you cannot summarise probably does not have a point yet.
This is also why writing feels hard when the thinking is unfinished. A paragraph you cannot open with a claim is usually a paragraph where you have not yet decided what you think — which is a thinking problem presenting as a writing problem.
A thesis is far too long to hold in mind at once. A reader — including an examiner — needs regular reminding of where they are and why this section exists.
What helps:
Signposting can be overdone — a paragraph of throat-clearing before each section is tiresome. A sentence or two is usually enough.
The most common practical problem is not writing badly but not writing at all, and it usually rests on a false belief: that you should understand something fully before writing about it.
In research the reverse is closer to true. Writing is how you find out what you think, and what you can defend. A section that felt clear in your head routinely falls apart on the page — and discovering that in month eight is far better than in month thirty.
What helps in practice:
Supervisor comments are frequently brief, sometimes blunt, and often describe a symptom rather than a cause. A few translations that recur:
If the same chapter keeps returning with comments about writing, the problem is usually the argument rather than the prose. Editing sentences will not fix a section that does not know what it is establishing. Ask your supervisor directly what the chapter is supposed to establish, and whether it does — a question that often resolves months of circling.
On editing help: proofreading and language editing are ordinary parts of academic work, and most universities publish guidance on what is acceptable — typically correcting language, clarity and consistency while leaving the substance and argument as yours. Check your institution’s policy, follow it, and acknowledge assistance where asked. It is straightforward and takes the question off the table.
Three changes fix most of it. Give each paragraph one point and state it in the first sentence. Turn nouns back into verbs — “introducing the intervention improved engagement” rather than “the implementation of the intervention resulted in an improvement”. And hedge once rather than four times: pick the confidence your evidence supports and state it. None of this means simplifying the ideas; technical vocabulary is precision.
Most often paragraph structure. If a paragraph contains several points and the main claim arrives in the middle, the reader has to assemble the argument themselves. Try reading only the first sentence of each paragraph in a chapter — together they should summarise the argument. If they do not, the paragraphs are not leading with their points, or the argument has no sequence.
It depends on your discipline, and conventions have relaxed considerably. Many fields now accept first person, particularly in qualitative research where the researcher’s position is part of the account, and some methodological traditions require it. Others still prefer impersonal constructions in methods sections. Look at recent theses passed in your own department — that answers it more reliably than any general rule.
Accept that you write in order to find out what you think, rather than the other way round. Separate drafting from editing, since doing both at once is why people rewrite the same sentence repeatedly — write badly on purpose and fix it later. Draft the methodology while decisions are fresh, keep a decisions log throughout, and write regularly in short sessions rather than in occasional long ones.
Usually one of four things: judge the quality of the evidence rather than reporting findings, attend to what each study’s design could actually establish, examine assumptions the field treats as given, or take a position where sources disagree. It does not mean finding fault with everything — a chapter dismissing each study in turn undermines the foundation you are about to build on.
Proofreading and language editing are ordinary parts of academic work, and most universities publish guidance on what is acceptable — typically correcting language, clarity and consistency while leaving substance and argument as yours. Check your own institution’s policy and follow it, including any request to acknowledge editorial assistance. Declaring it is straightforward and removes the question entirely.
If comments are about the writing but the same chapter keeps returning, the problem is usually the argument — and that is quicker to diagnose from outside. Send it and a PhD in your field will tell you what it is missing.
Leave an email or a WhatsApp number — whichever you prefer — and tick how we should reach you. We reach out within 30 minutes.
