Research
Objectives
The NSF project will substantially expand our knowledge of the maritime dynamics of the Spanish colonial empire and its primary viceroyalty, the New Spain (today’s Mexico). By applying state-of-the-art Artificial Intelligence methods, including:
- Handwritten Text Recognition
- Natural Language Processing
- Corpus Linguistics
- Geographic Information Systems
- Computer Vision techniques
The project will create a corpus of digital historical documents (transcribed and annotated) that would otherwise take generations to produce.
The NSF project draws on the team’s previous experience in successfully creating methods, tools and datasets and applying computational techniques to analyse sizeable historical document collections.
This work has included the study of some of the most important documents and maps of colonial Latin America and the creation of unparalleled digital resources that facilitate large-scale research on the inland geographies, toponyms, and landscapes of New Spain (see projects Digging into Early Colonial Mexico, Unlocking the Colonial Archive)
Questions
By applying AI methods in a Big Data scale to the study of the NSF, the project will answer the following methodological questions:
- • How can we advance AI methods for historical research?
- • How can large-scale approaches substantially facilitate the analysis of historical documents and the identification of information and patterns that otherwise would be impossible to find and study?
- • Through the Case Studies, the project will aim to answer specific historical questions as described in Research Area 3.
RESEARCH
AREA 1
Automated transcription
In this research area, we will classify the entire collection by calligraphy type and develop Handwritten Text Recognition (HTR) models for its automatic transcription. For this task, we will use Transkribus, a comprehensive HTR platform for the text recognition, image analysis and structure recognition of historical documents. It automatically converts text on an image to characters using machine-learning techniques, allowing the automated transcription of historical documents.
The NSF project will leverage the preliminary models being currently developed by the AHRC-funded Unlocking the Colonial Archive project, refining them by augmenting the training sample. In addition, the project will further expand the repertoire of automatic transcription models by training sets for six 16th – 18th -century calligraphies: cortesana, gótica, procesal, procesal encadenada, itálica cursiva and itálica redonda.
This Research Area will produce:
A machine-readable corpus of historical documents containing the transcriptions of thousands of ship records (the NSF Digital Corpus)
A collection of six HTR models for the automatic transcription of Spanish calligraphies ranging from the sixteenth to eighteenth centuries.
Creating an annotated corpus through Natural Language Processing (NLP) techniques opens the possibility of implementing text-mining approaches and carrying out the identification, cross-linkage, and analysis of information in thousands of documents. In this Research Area, we will create a machine-learning model for automatically annotating information of interest in the entire NSF Digital Corpus.
This model will consider part-of-speech as well as concepts and keywords, including semantic relationships between them. The team will use NLP technologies for automatically annotating documents, expanding the ontology and machine learning model developed by the ESRC-funded Digging into Early Colonial Mexico project.
The ontology, based on a Linked Open Data model, contains 18 different entities, each with several labels for annotating a wide variety of social, cultural, economic, geographic, natural and conceptual features, ranging from persons and locations, through means of transportation and measurements, to resources, plants, and animals among many others. Although this ontology is already quite comprehensive, the NSF project will expand it, including entities that have not been considered yet, particularly those that will enable the identification of keywords and concepts related to seafaring and maritime trade.
This Research Area will produce:
An NLP model for the automatic annotation of entities and semantic relationships of information in transcribed historical sources that will enable the identification of a variety of information, including maritime topics
The NSF Annotated Corpus that will serve as the basis for case studies
An Online Platform where the Corpus can be consulted as an open-source and open-data development.
RESEARCH
AREA 2
Automated annotation
RESEARCH
AREA 3
Case Studies
In this research area, the project will explore five case studies, answering specific questions that will yield new insights into the Transatlantic and Colonial history of the Americas, and showcase the benefits of working with AI approaches.
NSF Social Networks
We will aim to reconstruct the social network of an entire New Spain fleet to explore the fleet’s connections among individuals and social groups.
Questions: What was the structure of social relationships in the NSF of 1630-1631? How did variables such as geographical origin, social background, traded merchandise, or professional expertise relate to how people connected? Which individuals or clusters had pivotal roles in the fleet’s overall social structure? How did the social networks of people travelling in the fleet extend beyond the limits of the fleet itself? How did important events such as the 1631 naval accident modify the fleet’s social network structure?
Trade routes
Integrative connectivity approaches to coastal and inland routes, which consider other potentially important locations and riverine and terrestrial routes, have only recently emerged. The identification of key coastal, riverine, and terrestrial trade nodes and pathways will provide an unprecedented understanding of Spanish colonial trade mobilities, which may be traced back to late Mesoamerican routes.
Questions: What were the key coastal, riverine, and terrestrial nodes and pathways of the NSF trade routes? Was there a hierarchy of transit nodes in the NSF trade routes? Did they change over time? How did coastal trade nodes connect with fluvial and inland trade nodes? Who were the most frequent users of such trade nodes, and what was the nature of their mobility?
Scientific knowledge, medicine, and exchange of indigenous plants
The recording, cataloguing, and exchanging of flora and fauna between previously distant worlds in the sixteenth century would transform knowledge, economies, tastes, and ecologies. Despite the many studies carried out in the last century, there are still many questions to explore regarding the economic, social, and scientific value of American plants.
Questions: Who was involved in exchanging these crucial resources? Which plants reached the old continent first, and why were those prioritised? Which plants were more frequently traded during this period and why? How did indigenous medical knowledge travel through the NSF?
Outward marks
Historical documents often contain non-textual symbols that could yield important insights into their social context. Such is the case of the marcas de afuera (‘outward marks’), a very salient and interesting feature in the NSF ship’s records that has been largely overlooked in existing research. Analysing such non-textual symbols represents an ideal case study for joining economic and social history with iconographic studies and computer vision developments.
Questions: What were the iconographic characteristics of outward marks? Can we trace marcas de afuera (and therefore commodities) inland? How were marks associated with people or other elements of the NSF trade? How did the frequency or complexity of marks change over time? What was the relationship between outward marks’ use and changes in fleets’ composition or regulations?
Indigenous slave trade
Despite official banishment by the Spanish crown in the early 16th century, enslavement and trading of indigenous people persisted in the New Spain and took various forms across the Spanish colonial period. Evidence is emerging that indigenous people were sent to the Caribbean and Europe. However, little is known about the maritime trade, the merchant network, and the infrastructure that enabled such activities. In this case study, we will identify and analyse information about enslaved people, their destinations, and involved tradespeople and ships from the entire NSF Corpus.
Questions: In which conditions and status did indigenous people travel in the NSF? Can we identify a systematic indigenous slave trade through the NSF Fleets? What were the origins and destinations of transported indigenous enslaved people? Which individuals and ships were directly involved in the trade? How did it evolve throughout time? Was the Spanish slave trade restricted to its colonies or related to other slave trade systems, such as the Portuguese, Dutch or British?
This Research Area will produce:
A catalogue of outward marks containing high-resolution images of all identified marks and information about their associated tradespeople.
At least five publications (one per case study) in high-impact scientific journals, presented at international conferences, and made freely available through the project's online website.