Academic Tools

Determines who wrote what — not by guessing but by measuring. It integrates multiple independent stylometric streams: function word distribution, sentence rhythm, punctuation habits, vocabulary richness, syntactic preferences. Results are reported as likelihood distributions across candidate authors, never as point identifications. Attribution is a probability problem, not a certainty problem.

Can help you with: Authorship attribution, disputed texts, detecting ghostwriting, verifying authenticity, multi-author detection within single works, and any problem where the question is “who wrote this?”

Checks every citation in a bibliography against live sources and returns a verdict on each: Verified, Inaccurate, Unverifiable, or Likely Hallucinated. A citation is a claim — like any claim, it can be true, false, or unverifiable. Plausibility is not accuracy. LLM hallucinations are designed to look real: real author names, plausible journal titles, believable volume numbers. The only test is checking. This tool checks, using a three-level cascade — DOI lookup, title and author search, journal and volume search — before reaching any verdict.

Can help you with: Verifying bibliographies before submission, detecting hallucinated citations, identifying garbled reference details, distinguishing real-but-paywalled sources from fabricated ones, and any situation where you need to know whether your references actually exist.

Reconciles conflicting dates, reconstructs uncertain timelines, and orders events from fragmentary evidence. Handles calendar conversions (Julian, Gregorian, Islamic, Hebrew, regnal years), resolves contradictions between sources, and builds probabilistic timelines when exact dates are unavailable. Distinguishes between what is dated, what is datable, and what can only be placed in sequence.

Can help you with: Date reconciliation, calendar conversion, timeline construction from fragmentary evidence, resolving chronological contradictions between sources, and any problem where the order of events matters but the evidence is incomplete.

Traces the biography of a word or concept across time — from its etymological birth through its migrations across languages, disciplines, and centuries. Where the Semantic Cartographer maps a word’s neighbourhood at a given moment, Etymology maps its life across all moments: when it was born, who adopted it, how it changed, what it replaced, and what will eventually replace it.

Can help you with: Etymology, semantic history, conceptual genealogy, tracking how ideas migrate between disciplines, and any problem where understanding where a word came from illuminates what it actually means.

Maps citation networks, traces allusions, and reconstructs the web of influence between texts. Every text is woven from other texts — quotation, paraphrase, structural borrowing, silent allusion. Identifies these threads and maps the topology of influence: who read whom, who borrowed what, how ideas propagated through networks of texts across centuries.

Can help you with: Mapping citation networks, tracing allusions and structural borrowings, reconstructing influence topology, identifying silent quotation, and any problem where you need to see how texts are connected to each other.

Reads arguments as structures, not performances. Maps the premises, inference, and conclusion of any argument; surfaces unstated assumptions; identifies and names logical fallacies precisely; distinguishes formal from informal errors; and assesses the argument’s overall strength. The argument you think you are analysing is rarely the argument that was made. The Logic Auditor finds the actual argument first, then judges it. It wants the argument to succeed — that is why it finds what fails.

Can help you with: Identifying logical fallacies by name, mapping argument structure, surfacing hidden premises, distinguishing valid from sound arguments, assessing rhetorical moves that are not strictly logical, and salvaging structurally flawed arguments.

Reconstructs the family tree of texts — the stemma codicum — from the pattern of shared errors across manuscript witnesses. Every manuscript was copied from another, and every copyist introduced errors. The pattern of shared and unique errors reveals genealogical relationships between witnesses, identifies lost intermediaries, and determines which readings are closest to the original. Works with any textual tradition: classical, biblical, medieval, or modern.

Can help you with: Stemmatology, manuscript collation, variant analysis, identifying scribal errors, reconstructing lost exemplars, and any problem where multiple copies of a text disagree and you need to determine which is closest to the original.

Reconstructs the social networks of the past from fragmentary evidence. Identifies persons from partial references, maps relationships (kinship, patronage, correspondence, intellectual influence, enmity), detects possible identity merges, and builds network graphs from scattered attestations. Works from late antiquity to the present: Roman senators, medieval monks, Enlightenment correspondents, modern academic citation networks.

Can help you with: Identifying historical persons, reconstructing social networks, mapping patronage and influence, detecting identity merges from partial evidence, and any problem where understanding who knew whom changes what you understand about what happened.

Reads other people’s papers so you can assess them properly. A paper is an argument in disguise — strip the method section, the literature review, the hedged language, and what remains is a claim, an attempt to prove it, and a set of assumptions that make the proof seem stronger than it is. The Research Auditor identifies the central claim, assesses whether the methodology can support it, evaluates the evidence, locates unstated limitations, and names what would falsify the argument. Different from the Academic Supervisor, which reads your work: this reads the work you are reading.

Can help you with: Critically assessing academic papers, evaluating methodology, identifying overclaiming, finding unstated limitations, stress-testing research claims, and understanding what a paper can and cannot establish.

Analyses how texts persuade. Detects rhetorical figures (anaphora, chiasmus, antithesis, tricolon, and dozens more), maps the dispositio (the structural arrangement of argument), analyses the balance of ethos, pathos, and logos, and compares rhetorical profiles across authors or periods. Every argument has an architecture. Rhetorical Analysis makes it visible.

Can help you with: Rhetorical figure detection, argument structure analysis, comparing rhetorical styles across authors, speech analysis, persuasion anatomy, and any problem where you need to understand not just what a text says but how it works.

Goes far beyond the traditional concordance. Where a standard concordance tells you where a word appears, the Semantic Cartographer maps where a word lives — its conceptual neighbourhood, its competitors, the history of its migrations through meaning-space. Returns semantic neighbourhood maps, polysemy certificates, drift reports, and competitor analyses. Applies equally to ancient philosophical vocabulary and contemporary political language where meaning shifts in real time.

Can help you with: Semantic neighbourhood mapping, polysemy detection, meaning drift analysis, competitor analysis between near-synonyms, tracking vocabulary across corpora and centuries, and any problem where you need to understand not just where a word appears but how it lives.

Takes apart compilations — texts assembled from earlier sources, reorganised and reframed. Most ancient and medieval texts are compilations, and many modern works are too. Identifies the sources, maps which passages were selected and which omitted, analyses the reorganisation strategy, and reconstructs the compiler’s method and purpose.

Can help you with: Identifying sources within compilations, analysing selection and omission patterns, reconstructing compiler methods, understanding how encyclopaedias and anthologies were assembled, and any problem where a text is built from earlier texts.

Recovers lost authorial identities from textual traces. Identity is not stored in content but in structure — and structure survives. A lost Ciceronian text paraphrased by a ninth-century monk leaves behind morphosyntactic fossils, rhythmic cadences, rhetorical fingerprints, and patterns of conceptual association that no amount of rewriting can entirely suppress. Eight independent signal classes are brought to bear: lexical preferences, syntactic habits, prosodic rhythms, biographical-textual markers, intertextual citations, argumentative structures, conceptual co-occurrence topology, and rhetorical figure distribution. Attribution is always a probability distribution, never a point identification.

Can help you with: Attributing anonymous or pseudonymous texts, recovering lost authors from fragments embedded in later works, stylometric analysis, forensic linguistics, detecting ghostwriting, and any problem where you have a text and need to find — or rule out — a writer.

Takes two or more translations of the same source text and identifies where the translators diverged — and what their choices reveal about interpretation, ideology, audience, and the limits of equivalence. Works across any language pair and any period: Homer into English, Aristotle into Arabic, the Bible into everything. Every translation is an interpretation. This tool makes the interpretive choices visible.

Can help you with: Comparing translations, identifying translator choices, analysing semantic shift across languages, understanding what is lost and gained in translation, and any problem where the same text exists in multiple versions.

Plagiarism is not the real problem. The real problem is that most students do not understand what it means to use a source. The Academic Integrity Coach teaches honest source use through the act of using sources: it shows you the difference between decorative citation and genuine engagement, between paraphrase-as-comprehension and paraphrase-as-disguise, between acknowledging a source and actually reckoning with it.

Can help you with: Understanding when and how to cite, learning to paraphrase without plagiarising, distinguishing your voice from your sources, using sources as genuine intellectual partners, and developing the habits of honest academic writing.

Reads your draft as a senior academic supervisor reads a draft before submission — looking for what the argument actually says, not what you intended it to say. A good supervisor reads your work once to find what it is actually arguing, which is rarely what you think you are arguing. That gap is where the work begins. The Academic Supervisor assesses thesis strength, argument structure, evidence, literature engagement, methodology, and prose clarity. It does not write for you. It tells you what needs to change and why.

Can help you with: Pre-submission feedback on essays, dissertations, and papers; identifying thesis weaknesses; assessing argument structure; evaluating evidence quality; reviewing literature engagement; diagnosing prose problems; and getting an honest verdict before you submit.

Most people look at a text. Very few read it. Close reading means seeing what is actually on the page — not what you expected to find, not what the author intended, not what the argument requires. It means attending to the specific word chosen over the obvious one, the sentence that does more than it appears to do, the passage that says something other than what it seems to say. Show this coach a passage and it will show you what you missed.

Can help you with: Close reading technique, evidence extraction from primary texts, distinguishing observation from interpretation, finding competing interpretations in the same passage, and developing the habit of actually reading rather than merely scanning.

Most arguments fail not because they are wrong but because they are imprecise. The Critical Thinking Coach tests claims for validity, identifies hidden assumptions, locates where reasoning breaks down, and helps you state what you actually mean rather than what almost approximates it. It does not tell you that you are wrong — it shows you precisely where, and why, and what a cleaner formulation would look like.

Can help you with: Testing the validity of your arguments, identifying hidden premises, distinguishing what you can claim from what you cannot, improving precision of reasoning, and developing habits of rigorous intellectual expression.

A good essay is a well-built argument. Every paragraph is a unit of reasoning. Every sentence does a job. The Essay Structure Coach reads your structure — not your prose, not your ideas, your structure — and shows you where it holds and where it collapses. Where sections do not serve the argument. Where the architecture contradicts the thesis. Where the reader is lost not because the idea is difficult but because the building has no doors.

Can help you with: Argument architecture, paragraph construction, claim-evidence-analysis patterns, logical sequencing, identifying sections that do not serve the argument, and rebuilding essays that have the right ideas in the wrong order.

Applies the Socratic method — the elenctic protocol of definition, counterexample, revision, and repeated examination — to any argument, claim, or position you bring it. This is not the historical Socrates, constrained by fifth-century Athens; this is the method extracted from the person, applicable to any topic in any era. The Examiner does not lecture, evaluate, or return a verdict. It questions. The examination is the product. A strong argument passes. Either outcome is progress.

Can help you with: Stress-testing arguments before presentation, debate preparation, finding weak premises through questioning, cross-examination practice, understanding positions you disagree with at their strongest, and discovering what you actually believe when pressed.

The strongest version of your opponent’s argument is the one worth defeating. If you cannot state it better than they can, you do not yet understand it. This tool teaches intellectual engagement through the practice of steelmanning — finding and articulating the most defensible form of a position before attempting to refute it. Used for debate preparation, genuine disagreement, and the discipline of not arguing against positions nobody actually holds.

Can help you with: Constructing the strongest version of opposing arguments, preparing for debates, ensuring you are engaging with real positions rather than convenient ones, and developing the intellectual habit of understanding before refuting.

Finds and constructs the strongest possible version of any argument — regardless of whether you agree with the conclusion. The strongest version of an argument is not the version the person making it gave you. It is the version a brilliant, well-informed advocate would give if they had time to prepare. The Steelman Engine finds that version: upgrading the evidence, selecting the most appropriate framework, handling the obvious objections from within, and cleaning the rhetoric. It has no view on whether the argument is correct. Its function is to build, not to judge.

Can help you with: Building the strongest case for any position, debate preparation, understanding opposing views at their best, testing your own arguments against their steelmanned counterparts, and ensuring you engage with the real strength of a position rather than a convenient weakness.

Most students answer the question they wished had been asked rather than the question that was asked. The Task Analyser decodes what is actually being required: the difference between “discuss” and “evaluate”, between “compare” and “contrast”, between a question that asks for summary and a question that asks for argument. It identifies scope, unpacks instruction verbs, locates the implicit requirements, and shows you exactly what you need to produce before you produce it.

Can help you with: Decoding essay and exam questions, understanding what different instruction verbs require, identifying scope and boundaries, distinguishing what is being asked from what you assumed was being asked, and avoiding the very common error of answering the wrong question extremely well.