Lexis, innovation, diffusion, and frequency

Lexis and lexical innovation

Lexical innovation and diffusion

  • Society continually changes as new practices and products emerge (e.g. smartphones)

  • These changes typically first manifest themselves in language on the level of lexis in the form of neologisms (e.g. the words smartphone or iphone).

  • Which recent neologisms can you think of?

    • to google
    • staycation, baecation
    • medfluencer
    • byelingual
    • glamping
    • hangry
    • influencer
    • gender pay gap
    • kindergarchy
    • cringe
    • social distancing, frontline worker
    • millenium bug
    • to furlough, dotard

  • patterns

    • motivations
      • communicative need
      • describe a feeling/phenomenon
      • one-off
      • appeal
    • forms
      • blending

  • Knowledge of words is conventional: speakers learn form-meaning pairings.

  • Model of the Linguistic Sign (de Saussure 1916)

Theoretical framework

S-curve model

The S-curve model is relevant to linguistic innovation, diffusion, and language change.

Integration of Milroy’s and Rogers’ model of diffusion stages into an S-curve. (Kerremans 2015: p. 65)

The Entrenchment and Conventionalization Model (Schmid 2015, Schmid 2020)

  • The more frequently a word is used, the more likely it is
    • that speakers have stored it in their mental lexicon (entrenchment) and
    • that it is part of the conventional language system (conventionalization).
  • Usage, entrenchment, and conventionalization are interconnected.

(p. 4)

Operationalization

Frequency as an indicator for entrenchment and conventionality (Stefanowitsch and Flach 2017).

  • corpus-as-input: language used in corpora represents potential exposure to speakers
  • corpus-as-output: language used by speakers in corpora represents potential degrees of entrenchment

Pathways of diffusion

  • types of linguistic variation and diffusion

  • dimensions of diffusion
    • across speakers and communities
    • across text types
  • examples for different degrees of diffusion

Empirical analyses of diffusion based on frequency

Würschinger, Quirin. 2021. ‘Social Networks of Lexical Innovation. Investigating the Social Dynamics of Diffusion of Neologisms on Twitter’. Frontiers in Artificial Intelligence 4:106. https://doi.org/10.3389/frai.2021.648583.

Total frequency

Most frequent

Median

Least frequent

Cumulative frequency

Temporal dynamics of use

Volatility

Social dynamics of diffusion

Investigating frequency using Sketch Engine

Frequency over time

Using the English Trends (2014–today) corpus; example: blockchain

Timeline view

Frequency view

Frequency across text types

Using the enTenTen21 corpus; example: alt-right

Distribution across Topic text type

Practice: Lexical innovation

Use the above case study words:

  • upskill
  • hyperlocal
  • solopreneur
  • alt-right
  • alt-left
  • poppygate

Tasks

  1. In the English Trends (2014–today) corpus:
    • determine total frequency for each word
      • example for alt-right:
        • query: [lemma="alt-right"]
        • absolute total frequency: 95,094
        • relative total frequency: 0.96 per million words/tokens (pmw)
    • identify the year of highest usage
      • sort by year
      • use Frequency (raw) and Relative in text type (per million words)
  2. In the enTenTen21 corpus:
    • determine in which Genre each word was used most frequently
      • example for alt-right:
  3. Compare results:
    • Which words show highest/lowest degrees of conventionality?
    • For which words is there a discrepancy to the results on Twitter?