Open a paper in a field you know nothing about and the words that stop you are mostly not the field's words. They are constitute, derive, substantial, albeit, whereas: general vocabulary that turns up in every discipline and belongs to none of them.
The technical terms are the layer everyone worries about and the easiest layer to deal with, because they are defined in the text, listed in the glossary, and there are not many of them. The argument vocabulary is longer, invisible, and nobody hands you a list of it unless you go looking.
Eight words that do the work in academic prose
Academic English is not harder vocabulary; it is a different, small set of words used precisely and repeatedly.
- significant, meaning statistically real rather than large. This is the single most misused word in student writing.
- approach, a way of tackling a problem, and it is a noun far more often than a verb in this register.
- establish, to show something firmly, and stronger than show or find.
- demonstrate, to prove by evidence, and it takes a that clause: the data demonstrate that.
- indicate, weaker than demonstrate, meaning the evidence points that way without settling it. The pair is a hedge and a claim.
- consequently, therefore, marking a result, and it belongs at the start of a sentence.
- nevertheless, marking a contrast that survives what came before.
- furthermore is overused. It is correct and it is the clearest signal of a first draft written to fill a word count.
It has been counted, twice
Averil Coxhead published the Academic Word List in 1998: five hundred and seventy word families that appear across academic writing in every discipline and that are not in the first two thousand general families. Together they account for about a tenth of the words in an academic text, which sounds small until you notice where they fall. They are not decoration. They are the verbs and connectives holding the argument together, so a tenth of the words carries most of the structure.
Gardner and Davies rebuilt the count in 2014 from a much larger body of academic text, and the Academic Vocabulary List that came out of it reshuffles the membership without disturbing the finding: there is a general academic register, it is a few hundred families wide, and it is learnable in a term.
Where it stops: revising the glossary
The stall is doing the reachable half. A student revises the subject's terms, because they are listed, finite and already half known, and never touches the argument vocabulary at all. The result is a reader who can define every technical term in a chapter and cannot follow the paragraph that uses them, and who concludes the problem is the subject.
It survives because the diagnosis is wrong in a way nothing corrects. Failing to follow a paper feels like a gap in the field, not a gap in English, and the words that were actually doing the blocking are ones you have seen a thousand times and half know, so they do not present themselves as unknown. Nobody looks up constitute.
Finding out which band your unknowns are in
Lextutor's VocabProfile settles the diagnosis in about a minute. Paste in a page of something you actually have to read and it colours every word by frequency band, so you can see whether what is blocking you is the first two thousand, the academic set, or genuine technical vocabulary. It is free with no account and it replaced guessing which words in a reading are worth the effort, which is a guess almost everyone gets wrong in the same direction.
What it profiles is a text, not you. It will tell you that a page is mostly academic vocabulary; it cannot tell you which of those words you already have, so the output is a shortlist to check yourself against rather than a list to learn. Profile three pages of the same journal and the words that come back academic in all three are your actual syllabus.
What it looks like when it goes wrong
The recognisable failure is an essay written with the right words in the wrong frames: make a research, this essay will talk about, I want to explain that. Every academic word is present and the register is not, because these words carry fixed partners and a fixed level of formality that no vocabulary list records.
The second failure is treating the list as a set of synonyms for words you already have, so utilise replaces use everywhere and the writing gets heavier without getting more precise. The academic set is not a formal register for ordinary meanings. It is a set of words for talking about evidence, cause and qualification, and using one where a plain word would do is the clearest sign of somebody who has learned the list rather than the job.
Because these words come as families rather than as single items, the endings do a large amount of the work here, which is one root, and the nouns, verbs and adjectives around it, and the fixed partners above are which words English puts together, and which it never does. Both sit under building the word count you need to say what you mean.
Knowing the words is not the same as being able to make the moves they exist for, and the most common of those moves is stating a claim without owning it, which is a grammatical choice rather than a lexical one: when what was done matters more than who did it.