It is gone midnight, the discussion chapter is nearly finished, and the last section you have to write is the one that admits everything your linguistics or literature thesis could not do. A vague, apologetic paragraph here does real damage: it either undersells three years of careful analysis or hands an examiner a list of objections you handed them yourself. Leaving it unwritten costs you a viva where the examiner asks the questions you should have answered on the page — and a rushed, generic version costs you almost as much, because “more research is needed” tells an examiner nothing about whether you understood your own thesis’s actual boundaries.
Why a generic limitations section fails a linguistics or literature examiner specifically
Limitations, delimitations and recommendations are three different things, and treating them as one paragraph is the first mistake. A delimitation is a boundary you chose — the corpus you selected, the period, the authors, the specific dialect or genre — and stating it is not an apology, it is a definition of your object of study. A limitation is a constraint you did not choose but that affects what your findings can support — a corpus size that limits how confidently a pattern generalises, a single-researcher coding process in a discourse analysis with no second coder, translated source material where nuance may not survive translation. A recommendation follows directly from a limitation: if the corpus was limited to one register, the recommendation is not “future research should study more text” in the abstract, but a specific next study that would address that exact gap.
A worked illustration, invented for this example and not a real submitted thesis: a candidate whose corpus-linguistics thesis analysed hedging devices in a corpus of 500,000 words of academic writing from one discipline would state, as a limitation, that the finding cannot be assumed to hold across disciplines with markedly different rhetorical conventions, and would state, as a delimitation, that the single-discipline focus was chosen precisely to control for disciplinary variation while establishing the pattern within one field first. The corresponding recommendation names the next step exactly: a comparative corpus study of the same hedging devices across two or three disciplines with contrasting rhetorical norms, not a vague call to “extend the study.”
Corpus-specific limitations in linguistics: size, representativeness, register
A linguistics thesis built on a corpus has a specific, well-known set of limitation categories to work through explicitly rather than gesture at. Corpus size limits statistical confidence in frequency-based claims — state the actual token count and what claims that size can and cannot support, rather than leaving the reader to guess whether your corpus was adequate for the kind of frequency comparison you ran. Representativeness limits how far a finding generalises beyond the specific register, genre, time period or speaker population the corpus was drawn from — a corpus of parliamentary transcripts tells you about parliamentary register, not spoken English generally, and a thesis that slides from the specific to the general in its conclusion without flagging that move is exactly what an examiner will stop you on at viva. Where a corpus was built from a single source type or a narrow time window for practical reasons, state that as a delimitation in the methodology chapter and revisit its consequences explicitly in the limitations section, rather than raising it for the first time at the end. The same sample-and-scope reasoning applies whether your evidence is a linguistic corpus or an interview sample — see how a political science thesis justifies its case or sample selection explicitly for the equivalent argument in a different evidence type, since an examiner in either field is checking the same thing: that the scope claimed in the conclusion matches the scope the evidence can actually support.
Literary-critical limitations: interpretive scope, translation, and the texts you did not include
A literature thesis’s limitations are usually about interpretive scope rather than sample size: which texts, editions, or authors were excluded from the analysis, and why that exclusion does not undermine the argument being made about the texts that were included. State explicitly whether you worked from original-language texts or translations, and if translations, name which ones and acknowledge that a translator’s choices are themselves an interpretive layer between you and the source — a genuinely common and defensible limitation in comparative or world-literature theses, provided it is named rather than left implicit. Where your reading draws on one critical or theoretical tradition rather than a plausible alternative — a psychoanalytic reading rather than a historicist one, for instance — state that as a delimitation of the interpretive lens chosen, and note briefly what a different lens might have surfaced instead, without attempting the alternative reading yourself in the same thesis.
State the edition used for every primary text as a specific bibliographic fact, not an afterthought in a footnote — different editions of the same literary work can carry meaningfully different texts, introductions and critical apparatus, and a thesis that treats “the novel” as a single stable object without naming which edition was read is leaving a genuine methodological question unanswered. Where a canonical work exists in a heavily variant textual tradition, state explicitly which edition or textual variant your analysis is built on and why, since a claim about a specific passage can simply be wrong if a different edition orders or words that passage differently.
Discourse and qualitative-coding limitations: reflexivity and the single-coder problem
Where the methodology involves qualitative coding of discourse, interview or textual data, the most commonly missed limitation is the single-researcher coding process: most doctoral discourse analyses are coded by one person, without a second coder for inter-rater reliability, and this is a legitimate, common limitation rather than a fatal flaw — but only if stated as such, with the mitigation you did use named explicitly (a coding framework agreed with your supervisor, a subset re-coded after a gap to check your own consistency, an audit trail of coding decisions). Reflexivity belongs here too: your own position relative to the material — a native versus non-native speaker of the language studied, a personal relationship to the literary tradition or discourse community analysed — shapes the reading you produced, and naming that position is a limitation statement, not a confession that weakens the thesis. Report how you addressed your own consistency directly: a coding framework agreed and documented with your supervisor before analysis began, a portion of the corpus or transcript set re-coded after an interval to check whether your own categorisation held up, or an audit trail showing how each code was defined and applied. A methodology chapter that names this mitigation is answering, before it is asked, the specific viva question about how a single coder’s interpretation was kept consistent across a large body of material.
Delimitations belong in the introduction, not only the limitations section
A delimitation stated for the first time in the limitations section, pages from the end of the thesis, reads as an afterthought rather than a deliberate scoping decision, even when the underlying choice was entirely sound. State your delimitations where you first define your object of study — in the introduction or methodology chapter — so a reader understands the scope of your claim from the outset, and then return to them briefly in the limitations section only to note any consequence of that scope that is worth flagging explicitly. A thesis that scopes itself to twentieth-century Anglophone poetry, for instance, should say so plainly in the introduction, framed as a deliberate focus rather than an implicit boundary the reader has to infer from which texts happen to appear in the bibliography.
Writing recommendations that examiners actually credit
A recommendation earns credit when it follows directly and specifically from a named limitation, not when it restates the thesis topic with “more research should be done” appended. Pair each limitation with one recommendation that a future study — or your own next project — could concretely execute: if the corpus was limited to written registers, recommend a specific comparative study incorporating spoken data of a named type; if the reading relied on one theoretical lens, recommend the specific alternative framework a follow-up study should apply and to what texts. This is the same logic a discussion chapter’s evidence-to-claim discipline requires more broadly — see the seven-move structure for a PhD discussion chapter for where the limitations section sits within that larger chapter, and how the moves that precede it set up what the limitations section then has to qualify honestly.
FAQ
What is the difference between a limitation and a delimitation?
A delimitation is a boundary you chose deliberately — your corpus, period, or texts; a limitation is a constraint you did not choose that affects what your findings can support, such as corpus size or a single-coder process.
Should I apologise for my thesis’s limitations?
No. State them factually and specifically, paired with what they mean for your findings’ scope — an apologetic tone reads as uncertainty about your own contribution, which is worse than the limitation itself.
Is working from translated texts a fatal limitation in a comparative literature thesis?
No, provided you name which translations you used and acknowledge the translator’s interpretive layer explicitly, rather than presenting a translated text as equivalent to the original.
Do I need a second coder for my discourse analysis to be valid?
Not necessarily — a single-coder process is common in doctoral research, but state it as a limitation and name the mitigation you used, such as a coding framework or a self-consistency check.
How specific should my recommendations for future research be?
Specific enough that a reader could design the next study from your recommendation alone — name the data type, method or framework, rather than a general call for “more research.”
Should limitations be discussed only at the end of the thesis?
Delimitations should be stated where you first define your scope, typically in the methodology chapter; limitations are usually gathered in a dedicated section near the end, but should not be raised there for the first time if they affect the methodology chapter’s design decisions.
Can too many stated limitations weaken my thesis?
A long, unfocused list can look like an examiner should have flagged them earlier — prioritise the limitations that genuinely bear on your central claims, rather than listing every conceivable constraint exhaustively.
Does a corpus size limitation apply to a purely literary-critical thesis with no corpus?
Not in the same statistical sense, but the equivalent question still applies: which texts, editions or authors were included and excluded, and why that selection does not undermine the interpretive claim being made about the material studied.
Write the section that closes the gap before your examiner opens it
The limitations section is where a thesis says, in its own words, exactly what an examiner would otherwise have to say for you — and doing that well, at 2am with a submission deadline close, is exactly the kind of section that benefits from a structured draft rather than a blank page. Tesify’s thesis writing workspace includes a free tier for drafting sections like this against your own corpus, texts and methodology, without your unpublished analysis or data ever leaving a workspace built for exactly this kind of confidential, in-progress research writing.