Brumberger: "The Rhetoric of Typography: The Persona of Typeface and Text"

Brumberger, E. R. (2003). The Rhetoric of Typography: The Persona of Typeface and Text. Technical Communication, 50(2), 206-223.





Possibilities afforded to writers via word processing/desktop publishing regarding design templates/ fonts.


Reshaping technical communication


roles of tech writers changed to where tc is both content and formatting choices.


Visual rhetoric and typography


TC research has looked at viz rhet but not typography. Cites Wheildon (1996) that typography rules = "craftlore." Most research focuses on readability/legibility.


The purpose of this project


Need to shift unconscious typography choices to conscious ones.


Related Literature


history of typographical analysis (e.g., serif [symbols of empire] vs sans serif [symbols of the republic]).


National/cultural identification of typefaces (Garamond=French, Bodoni=Italy, Caslon=England).


1920s: attempt to create ''supra-national'' (value-free?) typefaces--even if not national-identified, still rhetorical. Bauhaus: content and purpose of the text should dictate the design and form of the document.


Typeface persona in theory


Cites Arnheim and Lanham-- active and purposeful problem-solving process when we perceive. Readers make active judgments as they look at the page.


Typeface persona in practice


Practitioners tend to agree with the theory that viz attributes of a doc have a "subtle and often complex impact extending beyond legibility and readability" (208). Typeface "persona" often attributed to physical characteristics -- e.g., sans serif cleaner, more technical, less cluttered. Shape and weight an issue (rounded serifs "friendly" and squared serifs more "official").


Little empirical evidence of this, mostly considered intuitive, aesthetic judgment.


Typeface persona in research


Little research-- earliest in 1920s; most recent in 1980s.


atmosphere values: cheapness, dignity, economy, luxury, and strengths (1923); luxury/refinement, economy/precision, and strength (1969 and 1968)


professionals and laypersons: do they share same values? Brinton (1961) found general consensus between groups; Tannenbaum, Jacobson, and Norris (1964) showed relatively high agreement


semantic quality: Rowe (1982) potency [strong/weak], evaluation [clean/dirty], elegance, novelty, and antiquity. Bartram (1982) potency, evaluation, mood, and activity.


Need for additional studies


little consistency among results/personality profiles. no attempt made to establish personality profiles for both typefaces and texts


Research Methodology


2 empirical studies:


first, "whether readers/viewers consistently assign particular personality attributes to particular typefaces"


second "used comparable methodology to investigate whether they similarly attribute personality characteristics to passages of text" (209)




undergrad into psych students required to participates in four hours of research studies. no overlap between studies. 50/50 m/f. collected other demographx as well.


Data collection method


modification of semantic differential (Osgood, Suci, Tannenbaum 1957) (modification = using non-paired attributes). Avoided references to physical characteristics of the fonts.  Needed to be connotative. No explanations or descriptions to avoid bias.


Study 1: Establish Typeface Personas


Each of 15 typefaces, including Times New Roman, Garamond, Arial, and Novelty fonts. All displayed same size (14 pt).


Materials and procedure


packet given to 80 participants. Pages randomly sequenced. Administered in groups of up to 8. Brumberger able to observe during study.


Methods of Analysis


Used SPSS.


  1. analyzed rating scores for each attribute and typeface to determine if correlations between typefaces
  2. conducted exploratory factor analysis to identify common underlying factors in the pattern of ratings
  3. subjected data to multidimensional scaling (MDS) to uncover underlying structure of data
  4. analyzed to identify any f/x linked to age, major, usage/familiarity WWW, ethnicity, gendre




used means to determine correlations (high positive meant perceived to have similar personas; high negative means very dissimlar; highest possible is correlation of 1). Also computed significance


Factor analysis: Principle Components Analysis with Varimax Rotation (Kaiser Normalization). Sorting data until clearest groups are identified.


Multidimensional scaling: measures similarity/difference to generate spatial map of data




did "consistently ascribe particular personality attributes to a given typeface" (213).


Typeface persona: highly significant strong correlations.


Demographic differences: no significant effect, though infeasible to calculate age as only 7 of the 80 were not 18-23. Such varied groups no significant chunks of majors. Some intriguing observations regarding White and Hispanic participant reactions (216) and gender reactions (216)-- only notes "approached significance" (216).




shows evidence that ppl "consistently ascribe particular personality attributes to a particular typeface" (217).  Did not correspond to previous studies but may be due to diff methodology and passage of time.Study 2: Establishing Text Personas


use a text and analyze its "persona." (explains system for developing text samples). Printed in 12 pt Times New Roman.


Assessing reading comprehension and rate


assessed reading comp and rate. Used Nelson-Denny Reading test.


Materials and procedure


Same sample size (40/40 m/f) 15 text passages. Each ranked text on attributes. Brumberger there to observe.


Methods of analysis


used similar methods: determined correlations, conducted factor analysis; applied MDS; finally, secondary analysis to identify demografx and possible influence.




Again, consistently "ascribe particular personality attributes to a given text passage" (219).


Text passage persona: correlations highly significant. She did derive labels conducting an "informal analysis of the texts within each grouping"


Demographic differences: possible affects of gender and reading comp level. Again, not feasible to identify f/x of age.




strong evidence-- personality attributes to text. may be shaped significantly by demografx. suggest readers of different abilities interact with texts differently.




"The data from Studies 1 and 2 provided strong evidence that readers do consistently ascribe particular personality attributes to particular typefaces and text passages. The typefaces and texts used in the project separated into clear categories according to their personas, and the differences between the categories were substantial. The Study 1 results thus provide strong support for the speculative body of literature that argues that typefaces have personas. Additionally, the data supports theoretical perspectives that suggest that visual language is analogous to verbal language in carrying connotations" (221).


Implications and Directions for Future Inquiry


doc design is problem-solving task


potential importance of vizrhet


now, for tech comm, what is the why? important to understand the reasoning


participants were limitations due to sample-- perhaps conduct with both laypersons and design experts.