Students usually compose adequate descriptive abstracts, but many confuse summary abstracts with short paraphrases or descriptive abstracts. Textbooks define a summary abstract ambiguously, as a “mini-paper” and/or as a mere statement of an article's topic and conclusions; most textbooks maintain the conceptual distinction between summary and descriptive abstracts even though differences between the two types are blurred in practice. These irregularities are accounted for by a hypothesis: in all levels of discourse, from sentences to extended texts, general and specific components conserve the “shape” of information. Intermediate discourse components (e.g., sentential tense, the syllogistic middle term, or the body of a text) may be deleted to create a smaller equivalent discourse structure. The two polar abstract types represent polar (general vs. specific) text components. Common abstracting errors arise from two sources: failure to distinguish between an abstract as “mini-paper” and a short paraphrase from the body of a long text, but also failure to distinguish between general topical information and the specific claims of a text, attributed to students' usual lack of acquaintance with other literature on a topic, besides the article they attempt to abstract.

Common technical graphics terms table, graph, chart, and diagram share a parallel logical structure with the four common types of technical graphics that the terms typically refer to. In the system of terminology as in the system of graphics types, four logical categories result from the possible permutations of two features. The abstract semantic features which underlie the meanings of the terms are in this discussion labeled as [UNITS] and [PROPERTIES]; likewise the significant features which distinguish the graphics types are here labeled as “units” and “properties.” These proposed semantic features reflect a fundamental semantic relation common to all meaningful statements, the attribution of a property (a predicate) to an object (a subject). The connection between term-features ([UNITS] and [PROPERTIES]) and type-features (“units” and “properties”) is a variable but systematic sense-reference relation. Consequently the terminology used to refer to the various graphics types varies systematically according to the markedness relationships among the terms. Principled explanations of the best uses for each graphics type follow from the proposed logical relations between them.

This article proposes a means of characterizing the difference between technical and literary writing, involving a theory of representation in which these distinct writing types are comparable to distinct types of visual representation. Any difference is only intelligible relative to a background of similarlity, but recent discussions of technical writing emphasize its similarity to literature and ignore significant differences. Distinct types of line drawings replicate the literary/technical contrast in a visual medium. This arises from two factors: 1) the way in which the drawing/text is perceived by the viewer/reader, as a substitute or as a standard; and 2) the predominant type of detail in the drawing/text, iterative or contrastive. Literature is most effective if perceived as a substitute for reality, predominated by iterative detail. Technical writing is most effective if perceived as a standard for evaluating reality, predominated by contrastive detail.