DICTION (Software)

DICTION is a dictionary-based language analysis program that analyzes the implied meaning of a text [Page 216]by searching it with the assistance of some 40 dictionaries or word lists. DICTION's corpus consists of 10,000 search words, none of which is duplicated in its search routines. DICTION provides an unusually comprehensive examination of a given passage by comparing it with a 25,000-item sample of contemporary discourse. Although the dictionary scores are individually interpretable, the program also creates five master variables: optimism (language endorsing some person, group, concept, or event or highlighting its positive entailments), activity (language featuring movement, change, the implementation of ideas, and the avoidance of inertia), realism (language describing tangible, immediate, and recognizable matters that affect people's everyday lives), commonality (language highlighting the agreed-on values of a group and rejecting idiosyncratic modes of engagement), and certainty (language indicating resoluteness, inflexibility, and completeness as well as a tendency to speak ex cathedra). Correlations among these master variables are largely nonexistent, thereby affording five independent examinations of each passage analyzed.

Figure 1 DICTION 6.0: Project View

Note: DICTION 6.0 operates on Windows XP, Vista, Macintosh, and Linux platforms.

The current version of the program, DICTION 6.0, was developed in Eclipse using Java 5.0 and operates on Windows XP and Vista, Linux, and Macintosh platforms. It processes 2,500 text files within 1 minute and produces both project output and alphanumerical files for subsequent statistical analysis (Figure 1). The program accepts PDF, RTF, Microsoft Word, and HTML input files for processing. Users can also supplement the program's builtin search features with up to 30 customized dictionaries of their own creation. DICTION 6.0 comes with a file management system so that users can group texts based on semantic commonalities, highlight specific textual regions for inclusion or exclusion, and identify (via color coding) different speakers or passage segments.

DICTION has been used to study political messages, media reportage, corporate annual reports, historical and literary documents, religious sermonizing, economic forecasting, medical documents, crisis communications, and (increasingly) websites and internet traffic. Unlike other programs, DICTION is largely “deductive” in that its dictionary structure has been conceptually derived and it compares all output with a normative data bank, thereby highlighting a given text's rhetorical distinctions and permitting immediate cross-comparisons with other DICTION-processed texts.