What is ADAR?
ADAR — Advanced Discourse Analysis Research — is an intelligent text analysis platform. You paste or type any piece of writing, and ADAR reads it the way a trained discourse analyst would: it identifies the cultural values, communicative strategies, and culturally specific language embedded in the text — then turns those findings into clear, scored results you can use in research, translation, or writing.
Unlike basic text tools that count words or measure sentiment, ADAR goes deeper. It reads how meaning is made — what cultural assumptions the text carries, how the author is trying to affect the reader, and what makes this text belong to a particular cultural tradition.
Running Your First Analysis
Getting results takes less than a minute. Here is the step-by-step process:
Go to the Analysis page
Click Analyse Text in the navigation bar, or go directly to /ultimate-analyze. You will see a large text input area.
Paste or type your text
Copy your text and paste it into the box. Aim for at least a full paragraph — the more text you provide, the richer and more reliable the analysis will be.
Set the Text Year (optional but recommended)
If your text was written in a specific year — say, a 1963 speech or a 1920s novel — enter that year. This tells ADAR to apply the right cultural context for that era, making the results more accurate.
Click Analyse
ADAR will process the text. This typically takes between 5 and 30 seconds depending on length. A progress indicator will show while it works.
Explore your results
Results appear in labelled panels — Framework Elements, Cohesion Analysis, Visualisations, and Statistical Validation. Each panel is self-contained and can be read independently.
Understanding Your Results — The Five Layers
Every ADAR analysis produces results across five distinct analytical layers. Think of these as five different lenses through which the same text is examined simultaneously. Here is what each one means:
Cultural Schemata
The big-picture cultural themes or "mental templates" the text activates — for example, a Hospitality Schema, a Justice Schema, or a Religious Authority Schema. These are the deepest values embedded in the writing.
Speech Acts
What the author is doing with their words — making a promise, issuing a command, expressing condemnation, declaring a fact. Every sentence does something beyond its literal meaning; ADAR identifies what.
Pragmemes
Recurring communicative strategies — patterns of how the author pursues their goal across the whole text. For example, repeatedly appealing to divine authority, or systematically building in-group identity.
Practs
The specific sentences or phrases where each strategy actually happens in the text. These are ADAR's most granular findings — the exact linguistic moves that carry out the broader strategies.
Culturemes
Culturally loaded words and phrases whose full meaning requires cultural background knowledge. Examples: Bismillah, promissory note (in a civil rights context), Inshallah. These are the hardest items to translate faithfully.
Cohesion (LCT)
How tightly the text holds together as a whole. ADAR measures this across six dimensions, from the repetition of key words to the density of domain-specific vocabulary. See Section 5 for details.
What each result item shows you
Every detected element — regardless of which layer it belongs to — gives you three pieces of information:
- Name — the label for what ADAR detected (e.g., "Liberation / Freedom Schema")
- Score — how strongly present this element is in the text (shown as a bar and a number out of 100)
- Evidence — the actual sentence from your text that triggered the detection, so you can verify it immediately
Reading the Scores
Every detected element carries a score from 0 to 100. This is not a grade — it is a measure of how strongly that element is present in the text. Here is how to read it:
Strongly present. This element is clearly and definitively embedded in the text. You can use it with confidence in your analysis or research.
Present, with some ambiguity. The element is detectable but not dominant. Worth noting, but check the evidence sentence to make sure it fits your interpretation.
Weakly or marginally present. A possible trace, but treat it with caution. Not suitable for primary conclusions without further reading.
The Evidence Sentence
The evidence sentence is the most important part of every result. It is the actual passage from your text that activated the detection. Always read it. If it makes sense to you as an analyst — if you can see why that sentence would trigger that category — the finding is solid. If it surprises you, that surprise is itself an interesting analytical moment worth exploring.
Cohesion Analysis (LCT)
The Lexical Cohesion panel tells you how tightly the text holds together as a whole. It measures six different ways a text creates internal unity — from repeated keywords to the density of domain-specific vocabulary — and combines them into a single Overall Cohesion Score.
What the overall score means
Exceptionally tight, publication-standard cohesion. The text is structurally unified across all dimensions.
Strong cohesion — the level of a well-written academic or professional text.
Moderate cohesion. The text holds together but some sections may feel loosely connected.
Weak cohesion. The text may feel fragmented or informal. Common in short texts, transcribed speech, or early drafts.
The six dimensions at a glance
- Reference — how well pronouns, connectors, and linking words bind sentences together
- Repetition — how often key words recur, creating thematic continuity
- Lexical Phrases — the density of multi-word formulae that signal academic or professional register
- Collocation — how frequently meaningful word pairs appear close together
- Semantic Field — how concentrated the vocabulary is within a specific topic domain
- Integration — how all five dimensions work together as a unified system
Downloading Your Reports
After analysis completes, click the Download Report button to save a formatted document. Three report types are available:
Comprehensive PDF
Full academic report with tables, charts, and all findings. Best for research and documentation.
AFA Report
Focused on the five-layer framework analysis with evidence lists. Best for discourse analysis papers.
CSV Export
Raw data table of all elements for further statistical work in Excel, R, or SPSS. Available on Advanced and Ultimate plans.
Your analysis history
All your past analyses are saved in your Dashboard. You can return to any previous analysis, download its report again, or compare results across different texts at any time.
Translation Analysis
ADAR includes two specialised tools for translators and translation researchers:
Translation Quality Check
Compare a source text and its translation side by side. ADAR analyses both texts independently and then measures how well the key cultural elements — schemata, speech acts, and culturemes — have been carried across from source to target.
The result is a set of preservation scores and a list of what was lost in translation — specifically which culturemes, schemas, or speech act patterns are present in the source but absent from the target. This gives translators a precise, evidence-based checklist for revision.
Translation Strategy Recommendation (Skopos)
Submit a source text and ADAR will analyse its cultural density, register, and dominant schema, then recommend a translation strategy — for example: "Foreignising translation retaining culturemes verbatim with cultural annotations" or "Communicative translation prioritising target-culture equivalence."
This recommendation is grounded in Vermeer's Skopos theory and Al Sharoufi's cultural density framework — not guesswork.
Worked Example — King's "I Have a Dream"
To show what ADAR can do on a well-known text, here is a summary of findings from analysing Martin Luther King Jr.'s famous 1963 speech. This illustrates the kind of insight ADAR generates — and how to use it.
"I Have a Dream" — Martin Luther King Jr., Washington D.C., 28 August 1963
1,667 words · Text year: 1963 · 23 paragraphs analysed
Cultural Schemata — What ADAR found
8 schemata were detected. The two strongest — Liberation / Freedom Schema (score: 96) and Constitutional Promise Schema (score: 91) — appear almost equally throughout the speech. At the same time, the Racial Injustice Schema (score: 89) runs just as strongly. ADAR's finding: King sustains an unresolved tension between America's founding ideals and its racial reality across the entire speech — this is not a coincidence but the architectural spine of the argument.
Speech Acts — What the speech is doing
The dominant speech act is Commissive — King is making promises and prophetic declarations, not primarily issuing commands. This is why the speech creates solidarity rather than merely mobilising action: King positions himself as a prophet announcing an inevitable future, not a leader demanding immediate compliance.
Pragmemes — King's key strategies
Six strategies were identified. The most powerful: the Anaphoric Insistence Strategy — "I have a dream" repeated 8 times, "Let freedom ring" 9 times, "Now is the time" 4 times. ADAR confirms what rhetorical scholars have long sensed: repetition is not ornamentation here — it is the primary structural strategy.
Culturemes — Culture-specific language
Six culturemes were detected, including "I have a dream" itself — now so culturally saturated it functions as a symbol beyond its words — and "promissory note," which frames civil rights as a debt America owes its Black citizens. These items are the hardest to translate without cultural loss.
Cohesion Score — 82.4 / 100
A 1,667-word political speech achieving the cohesion density of advanced academic prose. The dominant cohesive device is repetition — not the reference chains or academic vocabulary typical of journal articles. This confirms what King's speech is generically: oral prophetic literature, not conventional political oratory.
Tips for Getting the Best Results
- Use complete, flowing text. Bullet points, tables, or isolated sentences will not give ADAR enough linguistic context to detect patterns. Prose paragraphs work best.
- Always enter the Text Year when analysing historical texts. A speech from 1920 and a speech from 2020 carry different culturemes; telling ADAR the year ensures it uses the right cultural context.
- Read the evidence sentences. Every finding comes with a direct quote from your text. Always read it — it is your built-in verification mechanism.
- Focus on scores above 60 for your primary analysis. Lower scores are useful as supplementary or background context but should not anchor your main conclusions.
- Compare texts deliberately. ADAR is most powerful when you use it to compare two texts of the same type — two translations of the same source, two speeches from the same period, two news articles on the same event. The differences in layer profiles are where the real insights emerge.
- Use the Cohesion panel for writing feedback. If you submit your own draft and the Repetition score is very high relative to the other dimensions, your writing may be over-relying on keyword repetition and would benefit from more varied cohesive strategies.