Early research, such as studies published by Poffenberger and Franken (1923) and Schiller (1935), was primarily aimed at determining which typefaces could best be used to represent certain commodities in advertising. People were asked to rank a select number of typefaces with regard to the extent in which they were appropriate to commodities such as automobiles and coffee. Likewise, the subjects were asked to make a ranking with regard to qualities such as cheapness and dignity. Subsequent research has given increasingly more attention to the latter way of investigating the association of personality. The focus of studies seems to no longer be limited to advertising either. In addition, old terms, such as atmosphere value, have been replaced by new terms, connotative meaning.

The traditional term atmosphere value, also called feeling tone, is described by Ovink (1938) as "those properties by which [a typeface] excites feelings within the reader". In his extensive research of the phenomenon, Ovink not only found a collection of qualities like ‘unwieldy’ or ‘whimsical’, he also specified of both book- and display-typefaces, 30 in all, their status with regard to speed (liveliness and excitement), inclination (heartlessness and reserve), pastiness (sensuality and intellectualism), and many other associated properties. Analysis resulted in three general typeface clusters; namely ‘luxury-refinement’, ‘economy-precision’, and ‘strength’. While people are able to attribute characteristics to typefaces, he concludes that the average public will be unable to make the distinction between the atmosphere values of two typefaces which physically resemble each other. Consequently, the typographer "who did not hit upon the specially appropriate type, will not have done actual harm to the transmission of the meaning of the text, but ... missed an opportunity to intensify the force of impression of the text in a considerable degree" (Ovink, 1938).

More recently, the term congeniality is often used by both designers and researchers. The term refers to the "correspondence between content and visual form" (Zachrisson, 1965). In other words, congenial typography is the result of the successful use of a typeface, where atmosphere value and the actual content of the words set in that type share meaning. Harmony of form and content is exemplified by a love story printed in a delicate and warm typeface. If the atmosphere value of the type used would be one of rigidity and anger, the effect of that same story on the reader would be gravely disturbed, whether the reader would be consciously aware of this or not.

One problem in congeniality research is the impossibility of generalization. While research has been done using quite a number of typefaces, most studies have been carried out for use with a single application. For example, studying the appropriateness of Bodoni bold for the logotype of a computer company says nothing about the (im)possibility to use it on milk bottles. In an apparent attempt to stress that research can also be organized in another way, Rowe (1982) uses the general term connotative meaning rather than atmosphere value. She suggests that typeface connotation, preferred product connotation, and connotation of message content can all be assessed separately. This might enable us to use results of one particular study for more than one specific application by matching them with other results.

As a consequence of the need for results that are not limited to a single situation, focus of research has recently moved away from the purely commercial towards more theoretical areas. A more general or deeper understanding of the effects of congeniality on reading performance was the aim of studies by Bartram (1982), Walker, Smith and Livingston (1986), and Lewis and Walker (1989). Lewis and Walker use the term typographic allusion for atmosphere value, while Walker et al refer to the concept of multi-modal features. As with connotative meaning, it is supposed to be possible to determine the multi-modal features of typefaces separately from the multi-modal features of the message content. Walker et al (1986) asked subjects to indicate where particular professions should be placed on a range between features such as ‘soft and hard’, ‘weak and strong’, or ‘relaxed and tense’. The same was done with a number of typefaces and only then the professions and the typefaces with similar ratings were matched. In this way, typefaces with a known semantic quality can be used in a matchup for any application.

The recent studies just mentioned are in line with the argument that an instrument is needed to measure the general dimensions (factors) of typeface connotation. The need felt for standardization has even gone so far as to result in proposals of a classification of typefaces according to their atmosphere value. Although the scaling of values is generally considered useful, it is often argued that classification would be going too far (Zachrisson, 1965). One example of the development of such a universally applicable scale of values is Bartram’s (1982) research on semantic quality. Building on the results of previous efforts going back as far as Osgood, Suci and Tannenbaum (1957), Bartram found semantic dimensions that could be grouped into four factors: evaluation, potency, mood and vitality. Evaluation contains terms such as ‘pleasant’, whereas potency contains terms such as ‘strong’. Mood corresponds with ‘happiness’ and vitality implies ‘speed’, among other qualities. Bartram also found that designers as well as non-designers can agree to a significant degree on many typefaces studied with regard to at least one description of the four dimensions, but the groups often differ in judgment on other factors. This leads to the conclusion that typographers can not assume semantic quality in type based on their own judgment alone, but have to check which are the associations felt by the average reader. Using his own eye, which is more sophisticated, the typographer may present cues that the reader does not ‘understand’.

The work of Osgood et al (1957), mentioned before, has served as the basis for much research on the semantic quality of different kinds of objects, among which type. Based on this research, the Semantic Differential has been developed. With this method of data analysis, dimensions of connotative meaning can be determined. The results of investigations based on this instrument have brought us closer to the development of a practical instrument to determine typeface appropriateness for a specific application. In general, the results show that readers respond slower to text of which the form is inconsistent with the content. People also agree on the basic personality of a typeface up to a certain extent. Finally, typographers have a trained eye for typefaces and describe them with more adjectives than laymen. This allows typographers to make more subtle distinctions.

Ergonomics criteria as mentioned in the introduction are often not considered in research on type personality. Burt’s (1955) investigation of the relationship between legibility and reader preference of typeface is an exception. For different typefaces, the correlation between legibility and preference varied between 0.23 and 0.47. Tinker and Paterson (1942) also found that perceived legibility and pleasingness to readers are often in agreement. It should be noted, however, that preference does not necessarily indicate that content and form are in harmony; it may also mean that aesthetic appreciation partly determines ease of recognition. Ovink (1938) considered it proven that intense like or dislike of a typeface influences the speed of reading. In addition to this finding, it would also be interesting to investigate the relationship between personality and conspicuity. An example of a topic would be comparing the level of conspicuity of ‘aggressive’ typefaces with ‘intimate’ typefaces.

Finally, some general considerations to conclude the topic of the associated meaning of typefaces. It is important to note that association through convention is susceptible to change; type that has no distinct personality at one moment may well become typical of a particular institution later. For example, Copperplate Gothic is a typeface which has been used on displays of military material in the United States. Later, this typeface was also used on airplane displays in the assumption that it had proven its qualities. Consequently, this typeface proceeded from unassociated, through a typical military letter, to a typical airplane typeface.