How to use the Explorer — Cullen Project Explorer

What this tool does

The Cullen Project Explorer is an analytical companion to the Glasgow Cullen Project's scholarly edition of William Cullen's consultation correspondence (1755–1790). It provides tools for searching, comparing, and visualizing patterns across 5,600+ letters. The Explorer does not replace the Glasgow edition — every letter links back to the original transcription at cullenproject.ac.uk.

Building a subcorpus

Every analysis in the Explorer starts with a subcorpus — a filtered selection of letters. Go to the Build page and use the facets to define your selection: by voice (who wrote the letter), date range, corpus scope (full database or published file release), keywords (terms appearing in the letter text), concepts (tagged medical conditions, symptoms, or treatments), or person (a specific author or patient). For thematic exploration — such as disability, mental health, or pain — see the Research workbench, which provides curated word lists you can load directly into the builder. The live count updates as you change filters. When your subcorpus is ready, click one of the action buttons to take it into an analytical view.

Concordance — searching in context

Enter a search term to see every occurrence across your subcorpus, displayed in its surrounding context. Use the sort controls to cluster similar patterns: sorting by right context groups phrases like "pain in," "pain of," "pain was" together. Every result links to the original letter on the Glasgow edition website.

Keyness — comparing two subcorpora

Select two subcorpora (A and B) and the Explorer computes which words are statistically distinctive to each side. The register profile table below the keyword lists shows how linguistic features (pronoun rates, hedging, sentence length) differ between the two groups. Click any word in the keyness table to see its concordance within the relevant subcorpus.

Distribution — the shape of a subcorpus

See when your selected letters were written, who wrote them, and what conditions they discuss. Use this to check whether a subcorpus is skewed before running keyness, or to explore temporal trends in a topic.

Research workbench — exploring what the corpus contains

The Research page is the best starting point if you don't yet know what question to ask. It presents the full inventory of the corpus: every tagged condition, symptom, treatment, body part, ingredient, and therapeutic action, plus the people and places involved. Click any row to see its letter count and jump into concordance, keyness, or distribution analysis with that item pre-loaded.

Below the inventory, you'll find thematic word lists — curated sets of terms organised around research themes like disability, mental health, pain, fever, and women's health. These registers provide a ready-made entry point for specific scholarly questions and can be extended or combined in the subcorpus builder.

Medical glossary — understanding the terms

The Glossary page (under development) will provide definitions for eighteenth-century medical terms drawn from Cullen's own textbook, First Lines of the Practice of Physic. If you encounter an unfamiliar condition or treatment in the concordance results, the glossary will eventually be the place to look it up.

Showcase — pre-built analyses

The Showcase page presents curated analyses that demonstrate the tool's capabilities. Each showcase item tells a story with data and links into the relevant views so you can modify and extend the analysis.

Sharing your work

Every view encodes its full state in the URL. Click "Copy link" on any view to copy a shareable URL. Anyone who opens that link will see exactly the same analysis. You can also export concordance and keyness results as CSV files for use in other tools.

Tips for effective use

Start broad, then narrow. Begin with a large subcorpus and use keyness to discover what's distinctive, then drill into concordance to understand specific patterns.
Use the "Compare to rest of corpus" button on the builder to quickly see what makes any subcorpus distinctive.
The Research workbench's thematic word lists (disability, mental health, pain, etc.) are the best entry point for topic-driven analysis — they load directly into the builder's keyword filter.
Multi-word searches in concordance match exact phrases: "falling sickness" finds the phrase, not the individual words.
The file-release scope filter lets you restrict to the 3,798 letters in Glasgow's citable public data release.