Digital History Seminar, 19 Jan 2016
Dr Tessa Hauswedell
t.hauswedell@ucl.ac.uk
European or Imperial Metropolis? Depictions
of London in British Newspapers 1870-1900
Historical semantics
 Tracking changing meaning of terms over time,
space, and across different cultural contexts
 Challenge lies in tracking these changing meanings in
diachronic corpora.
Background to research
 Part of a European funded research project funded by
HERA.
 ‘Asymmetrical Encounters’ (asymenc.eu) uses digital
tools on 19th and 20th century newspaper corpora to
trace questions of transnational influences and role of
‘reference cultures’ in Europe between larger and
smaller West European countries.
Sources used
 EEBO (Early English Books Online)
 The Burney Collection: 17th and 18th Century
Newspaper Corpus
 British Newspaper Archive: 19th Century
 Pall Mall Gazette (full text archive from 1870-1900,
with thanks to British Library and J. Baker in
particular)
Tools and software used
 Antconc
 CQP (Corpus Query Processor)
 UCREL semantic tagset
All devised and maintained by Lancaster University,
freely available software
Methods
 Collocation Analysis provides insight into common
and frequent contexts within which a word is used.
 N-Gram or Cluster Analysis ( also called lexical
bundle) provides a sequence of words that form an
expression.
Paul Villars, London and its Environs. A picturesque survey of the metropolis and the
suburbs, p.5, 1888 Source: British Library Flickr Images
Illustrated London, or, a series of views in the British metropolis and its vicinity,
p.157. Source: British Library, Flickr Images
 ‘Metropolitanism’ as a term to describe the concerted
effort in 19th century to reshape relevant European
centers into dominant imperial/commercial centers.
 ‘the only cities that could afford such reshaping were
those that benefited from the colonial economy, the
metropoles of London, Paris and Vienna. Other colonial
capitals, like Amsterdam, Brussels, St. Petersburg, and
Berlin, and industrial ports, like New York, would follow
along as best they could, but always in the shadow of
these three metropolises’ (Rotenberg, ‘Metropolitanism
and the transformation of urban space’, American
Anthropologist, 2001, p.9)
 Claim that late nineteenth century discourse on
London took place within an ‘pan European
discourse’.
 Necessary to acknowledge the ‘distinctively European
dimension to the modern imperial city’.
 (Driver and Gilbert, ‘Capital and empire: geographies
of imperial London’, Geojournal, (2005) p. 23
Yorkshire Post and Leeds Intelligencer, December 12 1874, British Newspaper
Archive
Pall Mall Gazette, Maiden Tribute of Modern Babylon, July 1885
Distribution and dispersion breakdown for ‘metropolis’, Source: EEBO accessed via
CQP web processor
Collocations for ‘metropolis’ in EEBO corpus, 1600-1699, Source: EEBO corpus,
accessed via the CQPweb processor
Collocations for ‘metropolis’ in EEBO corpus, 1700-1799, Source: EEBO corpus,
accessed via the CQPweb processor
Source: 18th Burney Newspapers, accessed via Artemis Primary Sources
Source: 19th century British Newspapers, accessed via Artemis Primary
Sources
AntConc
 Cluster or N-gram Analysis
 Provides contiguous sequence
of words that form an
expression, can be any x-
number of n-grams
 Here between 2 and 5 n-grams
 Performed separately to the left
and to the right (before and after
the term) and for each decade
separately
 Collocation Analysis
 Provides terms of words that co-
occur frequently
 A span of +/- 5 words
 Collocation control can include
sentence boundaries
 Calculated on the basis of
Mutual Information Score
 Performed for each decade
separately
Mutual Information Score
 Common method to gain insight into actual strength of
the collocation
 Calculated on the basis of the number of times two
given words occur together versus number of times
they occur separately.
 The higher the score, the more likely that they are not
occurring together by chance.
 Can overestimate the score for collocations with very
low occurrences (less than 10).
Screenshot, Antconc
Clusters for ‘metropolis’, 1880-1989, PMGZ, arranged by MI score
Collocations for ‘metropolis’, 1890-1899, PMGZ, listed by MI score
UCREL semantic tagset
 Automated software tool for semantic annotation of
texts
 Includes 21 major discourse fields with subdivisions,
established as tagset for the semantic analysis of
terms in corpus linguistics
 Has 98 % lexical coverage for the 19th century,
through resources from EEBO as lexica
 Error rate of 8.95%
UCREL Semantic Analysis System
(USAS)
Source: http://ucrel.lancs.ac.uk/usas/
Source: http://ucrel.lancs.ac.uk/usas/semtags.txt
Examples of relevant discourse fields in relation to ‘metropolis’
Frequencies for search terms over period 1870-1899. Source: British
Newspaper Archive
Frames for the imperial
metropolis
 In a political context linked to discussions on the
British Empire
 In a national context relating to relations with Scotland
and or Ireland
 In relation to architectural schemes and infrastructure
projects
 In a historical context, as a comparison to cities of
classical antiquity
Conclusions
 From initially religious to more secular meaning and
increasing usage in public discourse throughout
nineteenth century.
 In the PMGZ, related to wide field of topical debates,
but no evidence of an implied European dimension in
relation to the metropolis, but instead to national
concerns.
 Overall London is presented as a metropolis quite
distinct from European counterparts in the late
nineteenth century.

European or Imperial Metropolis? Depictions of London in British Newspapers, 1870-1900

  • 1.
    Digital History Seminar,19 Jan 2016 Dr Tessa Hauswedell t.hauswedell@ucl.ac.uk
  • 2.
    European or ImperialMetropolis? Depictions of London in British Newspapers 1870-1900
  • 3.
    Historical semantics  Trackingchanging meaning of terms over time, space, and across different cultural contexts  Challenge lies in tracking these changing meanings in diachronic corpora.
  • 4.
    Background to research Part of a European funded research project funded by HERA.  ‘Asymmetrical Encounters’ (asymenc.eu) uses digital tools on 19th and 20th century newspaper corpora to trace questions of transnational influences and role of ‘reference cultures’ in Europe between larger and smaller West European countries.
  • 5.
    Sources used  EEBO(Early English Books Online)  The Burney Collection: 17th and 18th Century Newspaper Corpus  British Newspaper Archive: 19th Century  Pall Mall Gazette (full text archive from 1870-1900, with thanks to British Library and J. Baker in particular)
  • 6.
    Tools and softwareused  Antconc  CQP (Corpus Query Processor)  UCREL semantic tagset All devised and maintained by Lancaster University, freely available software
  • 7.
    Methods  Collocation Analysisprovides insight into common and frequent contexts within which a word is used.  N-Gram or Cluster Analysis ( also called lexical bundle) provides a sequence of words that form an expression.
  • 8.
    Paul Villars, Londonand its Environs. A picturesque survey of the metropolis and the suburbs, p.5, 1888 Source: British Library Flickr Images
  • 9.
    Illustrated London, or,a series of views in the British metropolis and its vicinity, p.157. Source: British Library, Flickr Images
  • 10.
     ‘Metropolitanism’ asa term to describe the concerted effort in 19th century to reshape relevant European centers into dominant imperial/commercial centers.  ‘the only cities that could afford such reshaping were those that benefited from the colonial economy, the metropoles of London, Paris and Vienna. Other colonial capitals, like Amsterdam, Brussels, St. Petersburg, and Berlin, and industrial ports, like New York, would follow along as best they could, but always in the shadow of these three metropolises’ (Rotenberg, ‘Metropolitanism and the transformation of urban space’, American Anthropologist, 2001, p.9)
  • 11.
     Claim thatlate nineteenth century discourse on London took place within an ‘pan European discourse’.  Necessary to acknowledge the ‘distinctively European dimension to the modern imperial city’.  (Driver and Gilbert, ‘Capital and empire: geographies of imperial London’, Geojournal, (2005) p. 23
  • 12.
    Yorkshire Post andLeeds Intelligencer, December 12 1874, British Newspaper Archive
  • 13.
    Pall Mall Gazette,Maiden Tribute of Modern Babylon, July 1885
  • 14.
    Distribution and dispersionbreakdown for ‘metropolis’, Source: EEBO accessed via CQP web processor
  • 15.
    Collocations for ‘metropolis’in EEBO corpus, 1600-1699, Source: EEBO corpus, accessed via the CQPweb processor
  • 16.
    Collocations for ‘metropolis’in EEBO corpus, 1700-1799, Source: EEBO corpus, accessed via the CQPweb processor
  • 17.
    Source: 18th BurneyNewspapers, accessed via Artemis Primary Sources
  • 18.
    Source: 19th centuryBritish Newspapers, accessed via Artemis Primary Sources
  • 19.
    AntConc  Cluster orN-gram Analysis  Provides contiguous sequence of words that form an expression, can be any x- number of n-grams  Here between 2 and 5 n-grams  Performed separately to the left and to the right (before and after the term) and for each decade separately  Collocation Analysis  Provides terms of words that co- occur frequently  A span of +/- 5 words  Collocation control can include sentence boundaries  Calculated on the basis of Mutual Information Score  Performed for each decade separately
  • 20.
    Mutual Information Score Common method to gain insight into actual strength of the collocation  Calculated on the basis of the number of times two given words occur together versus number of times they occur separately.  The higher the score, the more likely that they are not occurring together by chance.  Can overestimate the score for collocations with very low occurrences (less than 10).
  • 21.
  • 22.
    Clusters for ‘metropolis’,1880-1989, PMGZ, arranged by MI score
  • 23.
    Collocations for ‘metropolis’,1890-1899, PMGZ, listed by MI score
  • 24.
    UCREL semantic tagset Automated software tool for semantic annotation of texts  Includes 21 major discourse fields with subdivisions, established as tagset for the semantic analysis of terms in corpus linguistics  Has 98 % lexical coverage for the 19th century, through resources from EEBO as lexica  Error rate of 8.95%
  • 25.
    UCREL Semantic AnalysisSystem (USAS) Source: http://ucrel.lancs.ac.uk/usas/
  • 26.
  • 27.
    Examples of relevantdiscourse fields in relation to ‘metropolis’
  • 28.
    Frequencies for searchterms over period 1870-1899. Source: British Newspaper Archive
  • 29.
    Frames for theimperial metropolis  In a political context linked to discussions on the British Empire  In a national context relating to relations with Scotland and or Ireland  In relation to architectural schemes and infrastructure projects  In a historical context, as a comparison to cities of classical antiquity
  • 30.
    Conclusions  From initiallyreligious to more secular meaning and increasing usage in public discourse throughout nineteenth century.  In the PMGZ, related to wide field of topical debates, but no evidence of an implied European dimension in relation to the metropolis, but instead to national concerns.  Overall London is presented as a metropolis quite distinct from European counterparts in the late nineteenth century.