Home > Research > Publications & Outputs > Experiments in 17th century English
View graph of relations

Experiments in 17th century English: manual versus automatic conceptual history

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

Experiments in 17th century English: manual versus automatic conceptual history. / Pumfrey, Stephen; Rayson, Paul; Mariani, John.
In: Literary and Linguistic Computing, Vol. 27, No. 4, 2012, p. 395-408.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

APA

Vancouver

Pumfrey S, Rayson P, Mariani J. Experiments in 17th century English: manual versus automatic conceptual history. Literary and Linguistic Computing. 2012;27(4):395-408. doi: 10.1093/llc/fqs017

Author

Bibtex

@article{d681886b9ba24c4a8071d9388fb43f24,
title = "Experiments in 17th century English: manual versus automatic conceptual history",
abstract = "Previous research in conceptual history, the study of change over time of key terms and value systems, has been carried out manually using a restricted number of pre-identified texts. We propose that a method combining techniques from corpus and computational linguistics can be exploited to support conceptual history with semantic searches on a vast sample of texts. To exemplify this method, we focus on a fundamental concept in modern science, the experimental method, in order to trace how the pre-existing and primarily religious concept of experiment (or experience) took on its modern, scientific meaning. We contrast a manual approach using the existing Early English Books Online search interface with an automatic method using corpus linguistics software and methods to turn the transcribed portion of the same dataset into a corpus. Both approaches allow us to separate the religious and scientific senses and plot their change over time. We observe a rapid change in the meaning of experimental from overwhelmingly religious to largely scientific within the 1660s. However, the automatic corpus method is much more efficient and will support future scholars in carrying out iterative studies in a matter of minutes rather than through weeks of painstaking work. Such methodological innovation has the potential to support the formation of new research questions which could not have been considered previously.",
author = "Stephen Pumfrey and Paul Rayson and John Mariani",
year = "2012",
doi = "10.1093/llc/fqs017",
language = "English",
volume = "27",
pages = "395--408",
journal = "Literary and Linguistic Computing",
issn = "0268-1145",
publisher = "Oxford University Press",
number = "4",

}

RIS

TY - JOUR

T1 - Experiments in 17th century English

T2 - manual versus automatic conceptual history

AU - Pumfrey, Stephen

AU - Rayson, Paul

AU - Mariani, John

PY - 2012

Y1 - 2012

N2 - Previous research in conceptual history, the study of change over time of key terms and value systems, has been carried out manually using a restricted number of pre-identified texts. We propose that a method combining techniques from corpus and computational linguistics can be exploited to support conceptual history with semantic searches on a vast sample of texts. To exemplify this method, we focus on a fundamental concept in modern science, the experimental method, in order to trace how the pre-existing and primarily religious concept of experiment (or experience) took on its modern, scientific meaning. We contrast a manual approach using the existing Early English Books Online search interface with an automatic method using corpus linguistics software and methods to turn the transcribed portion of the same dataset into a corpus. Both approaches allow us to separate the religious and scientific senses and plot their change over time. We observe a rapid change in the meaning of experimental from overwhelmingly religious to largely scientific within the 1660s. However, the automatic corpus method is much more efficient and will support future scholars in carrying out iterative studies in a matter of minutes rather than through weeks of painstaking work. Such methodological innovation has the potential to support the formation of new research questions which could not have been considered previously.

AB - Previous research in conceptual history, the study of change over time of key terms and value systems, has been carried out manually using a restricted number of pre-identified texts. We propose that a method combining techniques from corpus and computational linguistics can be exploited to support conceptual history with semantic searches on a vast sample of texts. To exemplify this method, we focus on a fundamental concept in modern science, the experimental method, in order to trace how the pre-existing and primarily religious concept of experiment (or experience) took on its modern, scientific meaning. We contrast a manual approach using the existing Early English Books Online search interface with an automatic method using corpus linguistics software and methods to turn the transcribed portion of the same dataset into a corpus. Both approaches allow us to separate the religious and scientific senses and plot their change over time. We observe a rapid change in the meaning of experimental from overwhelmingly religious to largely scientific within the 1660s. However, the automatic corpus method is much more efficient and will support future scholars in carrying out iterative studies in a matter of minutes rather than through weeks of painstaking work. Such methodological innovation has the potential to support the formation of new research questions which could not have been considered previously.

U2 - 10.1093/llc/fqs017

DO - 10.1093/llc/fqs017

M3 - Journal article

VL - 27

SP - 395

EP - 408

JO - Literary and Linguistic Computing

JF - Literary and Linguistic Computing

SN - 0268-1145

IS - 4

ER -