Historical perspective

What are the origins of legibility research? A few landmarks are worth reporting as they formed the foundations for subsequent research and are frequently cited. Various writers (Spencer, 1968; Rehe, 1979; Tinker, 1965) propose that scientific legibility research began with Javal around 1880, a French ophthalmologist who studied eye movements in reading. In particular, Huey wrote a book on The psychology and pedagogy of reading in 1908 (reprinted 1968) which credits Javal with discovering the pattern of eye movements in reading (described in Chapter 2).

Closer examination of the original sources by psychologists reveals a different story (Wade, Tatler and Heller, 2003). In 1879, a physiologist Hering first discovered that reading involves saccadic eye movements and Lamare in 1892, a colleague of Javal, noted the jerky, rather than continuous, movements. Dodge (a psychologist) was able to develop a photographic technique which enabled more accurate measures of the speed of saccades and the length of fixations. These developments were the start of eye movement recording technologies.

Shortly after the discoveries concerning eye movements, Cattell (1886) claimed to have found the word superiority effect (as mentioned in Chapter 2). Other work around this time that is often cited includes:

• which letters are more legible than other letters (Sanford, 1888; Roethlein, 1912)

• a review of early legibility research by Pyke (1926)

Research directions

Although the above research looking at visual mechanisms in reading began in the late nineteenth century, visual science moved away from applied research and therefore lost a connection with legibility research. The psychology of reading became the province of cognitive psychology, education, and psycholinguistics with less interest in typographic and graphic aspects of text. From around 1980, computational models of reading were developed aiming to simulate the recognition of words through mathematical modelling with computers.

An extensive programme of legibility research was conducted by Tinker and his colleagues, which did not attempt to explain the underlying visual mechanisms for the results. To some extent, this reflected the state of knowledge at that time, and separation of theoretical and applied research. ‘Theoretical’, ‘basic’ or ‘pure’ research has the aim of investigating the visual processes involved in reading whereas applied research aims to evaluate which typographic solutions are better for reading. Tinker and colleagues carried out numerous experiments between the 1920s and 1950s which provide a substantial body of findings. Given our current knowledge of how we read, and more recent interest in visual processing relevant to legibility (Legge, 2007, p108), these results from traditional legibility research can now be more thoroughly evaluated and interpreted.

Reviews of legibility research summarise what is known at the time by discussing research published by others, and which might also include research by the author (e.g. Ovink, 1938; Tinker, 1963, 1965; Zachrisson, 1965; Spencer, 1968; Foster, 1980; Reynolds, 1984; Lund, 1999; Dyson, 2005; Beier, 2012). These can be useful texts for gaining an overview of research findings, which should again be critically evaluated.

Design perspective

Legibility research has not typically been carried out by designers, as they are unlikely to have a detailed knowledge of scientific method, normally acquired over the course of a psychology degree. However, designers have views on what should be researched and how it should be researched. These views do not necessarily fit with scientific approaches to research. The objectives of the research usually differ across disciplines and these can determine the methods used.

The primary method used by the psychologist concerned with reading research is empirical experimentation (Rayner and Pollatsek, 1989, p8). Within the psychology of reading there are various perspectives including cognitive and linguistic constructs (e.g. Kintsch and van Dijk, 1978), perceptual factors related to text legibility (e.g. Tinker, 1963, 1965), and the nature of visual information processing in reading (e.g. Legge, 2007).

A designer’s primary motivation is to make text more legible, i.e. easier to read, and also aesthetically pleasing — a practical approach. Designers often express dissatisfaction with certain aspects of empirical research, typically carried out by psychologists without consulting designers. Many of the reasons, raised in the past and still debated, are reviewed and summarised by Lund (1999) and Beier (2016). Below I list a few of these criticisms (from a design standpoint) of legibility research and then try to address these criticisms. These focus on the value and the relevance of the research to design practice. Issues which relate to the research methods are dealt with in the next chapter.

• A fundamental position opposing legibility research argues that research inhibits creativity, which seems to assume that the outcomes of research must be implemented. Poynor (1999, p14) suggests that designers view the scientific approach as an opponent of the creative process, and do not wish to be led by psychologists.

  • In 1970, Merald E. Wrolstad, editor of the Journal of Typographic Research speaking about the misunderstandings between disciplines, pointed out that research will never replace the creative designer; not everyone wants to be a typographer. Research should be regarded as a source of ideas (Wrolstad, 1970, p37–38). A solution may therefore be for designers to familiarise themselves with legibility research and then consider whether it is helpful to them in informing design practice. This does require researchers to present their work in an accessible form, and ideally including a hint as to how the findings might translate into practice, if relevant.

• Results of much legibility research merely confirm existing practices, based on craft knowledge. Or, results are contradictory and therefore of little use.

  • If empirical research supports existing practice, this can be seen as positive and reassuring. Where there are discrepancies, either between existing practice and research or among different studies, there is more cause for concern. Such issues may be resolved with further well-planned studies.

• Researchers naturally focus on areas and questions which interest them (Wrolstad, 1969, p116), which may not be particularly relevant to design practice.

  • Some research (such as the studies in Chapter 2) is not intended to have direct relevance to design practice. Pure research falls into this category. When reading about a study, it is important to work out why the research was done, which should be found in the introduction. The affiliations of the authors may also indicate their discipline if attached to a university or research centre and these are commonly included in the journal article (see Figure 3.1).

• Researchers should not look for universal truths but aim to provide practical decision-making guidance (MacDonald-Ross and Waller, 1975, p77).

  • Practical guidance may require testing a specific version of a design, which is better described as formative evaluation or user/usability testing, rather than legibility research. The results will apply to the particular situation and will not be generalizable. (See Chapter 4: Different types of testing and research).

• Researchers choose topics that are easy to investigate in the laboratory (MacDonald-Ross and Waller, 1975, p76)

  • This may have been true in 1975 and earlier, but I do not believe that this is currently the case as very sophisticated techniques have been developed. However, most of the research is done under controlled conditions (i.e. a ‘laboratory setting’). This is covered in more detail in Chapter 4.

Questions: If you were asked to debate the value of legibility research, which side would you prefer to argue: for or against?
Which of the points above do you think are the strongest?
Can you add any other points?

Combining resources across disciplines

Collaboration between people with diverse backgrounds and expertise can lead to mutual understanding of the important and different contribution that comes from another discipline. Engaging in discussion can help us understand the other’s viewpoint which should make us less dismissive of alternative perspectives.

Exercise:

Fernand Baudin (1918–2005), a Belgian book designer, author, typographer, and teacher, objected to Tinker’s description of typographers as aesthetes when reviewing the book ‘Bases for effective reading’ (Baudin, 1967). I have extracted excerpts from the pages listed by Baudin, which I think are the parts in the book that he references.

Consider whether you think Baudin was justified in being upset by the statements (quotations) below from Tinker’s book.

Do you think Baudin was right in interpreting the statements as: ‘…all typographers en bloc, whether expert or not, are presented merely as introspective aesthetes deserving, on the whole, of contempt’ (p204–205).

Is Tinker criticising typographers with these statements?
Is it an insult to be concerned with aesthetics?

• ‘Before scientific research, printers and type designers were concerned mainly with the esthetic appearance of the printed page.’ (p115)

• ‘…the dominant guides to typography until rather recently were esthetics, economy of printing, and traditional practice.’ (p125)
  ‘The subjective opinions of type designers and typographers as to legibility of letters prevailed throughout the nineteenth century and have carried much weight even up to the present day.’ (p125)

• ‘This practice continues even though many typography “experts” consider that italic type is far less legible than regular Roman lower case’. (p135)

• ‘Although some designers may have a strong esthetic objection to boldface for headings, this does not mean that readers react the same way.’ (p136)

• ‘The strong belief that generous margins will increase legibility agrees with the opinions of most “experts” expressed between 1883 and 1911 (Pyke, 1926) (p183)

• ‘While there is an “average” consensus, printing practice in use of margins in individual books varies greatly (Paterson and Tinker, 1940). Whether this is motivated by an attempt to produce a more pleasing page or by an unconscious departure from the 50 per cent rule, or both, is uncertain.’ (p183) [The 50 per cent rule refers to the general practice of publishers to use 50 per cent of a page for margins (Tinker, 1965, p182)].

If we look at the above quotations from a more neutral perspective, we might suggest that Tinker was wishing to make a clear distinction between scientific research (admittedly, his own) and the craft knowledge of typographic experts. The comments are not limited to aesthetics as legibility is included. However, it is unfortunate that Tinker uses quotation marks around the word “experts” which might be seen as an ironic comment.

Fortunately, we have moved on from Tinker and recognise that combining skills and knowledge across disciplines can result in more relevant and robust research. An example of an excellent collaboration between vision scientist and type designer is the article illustrated in Figure 3.1 and described in Panel 3.1. Other examples of collaborations where design expertise combines with scientific knowledge include:

• James Hartley and Peter Burnhill:
  Burnhill was a teacher of typography (1923—2007) who engaged in a long collaboration with James Hartley, a psychologist at Keele University, UK who is still a very active researcher and writer. This duo explored how typography can support readers’ use of texts through clearly displaying the structure of the texts (e.g. use of space).

• Robert A. Morris, Kathy Aquilante, Charles Bigelow, and Dean Yager:
  In 2002, vision scientists (Aquilante and Yager) combined with a mathematician working in computer science (Morris) and type designer (Bigelow) to look at how serifs affect reading on screen.

• Owen Churches, Scott Coussens, Hannah Keage, Mark Kohler, and Myra Thiessen:
  Thiessen is a designer with knowledge of how to conduct experiments; all the other members of the team are neuropsychologists and together they have looked at how the brain processes typography using EEG (electroencephalography) technology. Their research is mentioned in Panel 2.3.

Summary

Legibility research started with eye movement research over 100 years ago. Some of these discoveries and writings are still valid today, whilst others have been superseded as research has enabled more precise measurements and a larger body of knowledge has developed.

Legibility straddles disciplines, broadly science and design, and in the past, this has caused tension due to different objectives and, at times, insensitive appraisal of other perspectives. As more collaborations are developed, richer, more relevant, and more robust research findings emerge to inform typographic practice.