Project

Computational Models of Ambition in Early Modern Scotland

Overview

My doctoral research asked a deceptively simple question: why was ambition so universally condemned in sixteenth-century Scotland? To answer it, I built something unusual for a historian, a word-embedding model. By combining computational linguistics with archival research, I developed a method that could recover historically authentic meanings from centuries-old texts.

Challenge

Traditional methods stumble when faced with large, abstract concepts like ambition. Close reading gives nuance, but scaling across hundreds of texts risks flattening meaning or smuggling in modern assumptions. I needed a way to let the language of the period define itself, on its own terms, without imposing contemporary categories.

Approach

I created a 200-text corpus, prioritizing works documented in Scottish booksellers’ lists, wills, and libraries. Then I:

• Wrote custom Python preprocessing scripts to normalize archaic Scots and English spelling, tokenize, and lemmatize texts.
• Built a custom Scots dictionary for lemmatization, since no existing NLP toolkit supported the language.
• Applied Latent Semantic Analysis (LSA) to identify semantic clusters around “ambition” across the corpus.
• Validated digital outputs through close rhetorical reading and discourse analysis, situating computational results within cultural context.

This hybrid method let me trace how “ambition” was semantically linked to terms like pride, treason, power, and avarice, while also appearing in proximity to honour, virtue, and glory. The result: a historically grounded semantic map of ambition’s moral field.

Outcome

The research produced both a substantive historical argument and a methodological innovation:

• Historically: Ambition was condemned not merely as a private failing but as a systemic threat to legitimate authority. It was rhetorically linked to treason, rebellion, and social disorder, and controlled through mechanisms of law, religion, and reputation.
• Methodologically: I developed a replicable framework for applying computational models to humanistic questions, one that avoids anachronism and preserves historical nuance.

Reflection

This work demonstrated that computational methods can complement, not replace, traditional humanistic inquiry. The model became a bridge: allowing historians to interrogate concepts on their own terms while showing technologists how bias and exclusion are as old as language itself. After all: history may not repeat, but bias certainly does

Key Takeaways

• Built a 200-text historical corpus from 1500–1625 Scots and English sources.
• Developed a novel word-embedding methodology combining LSA with rhetorical analysis.
• Created custom NLP resources for early modern Scots.
• Demonstrated that ambition was semantically bound to treason, pride, and illegitimacy.
• Produced a replicable framework for computational historical linguistics.