The comprehensibility of several grammatical transformations.

Abstract

Sul Ross State College4 experiments compared the comprehensibility of different grammatical transformations of a passage. In 2 experiments, difficult prose was simplified by transforming nominalizations, adjectivalizations, and passives to their activeverb transforms. In the other 2, nominalizations alone were compared to their active-verb transforms. In all 4 experiments-which used several different presentation modes and several different dependent variables-the active-verb transforms were found to be more comprehensible.

The Effect of Transformational Simplification on the Reading Comprehension of Selected High School Students

“…Thus, there is a kind of serendipitous relationship between the de-transforming and recreating the sentencelength and internal structure apparently familiar to a population of 12th graders who read (and possibly write) at approximately 8th grade level. This relationship, between de-transforming and resultant sentence length, tends to corroborate the work of Coleman (1964) and others who have suggested these four transformations complicate the reading process because they encode information in sentence structures that are unused by many problem readers.…”

“…To begin with, the process of de-transforming the four problem structures identified by Coleman (1964) has manifold effects when it is applied to whole sections of running prose instead of isolated sentences. The first is that not every transformed structure may be de-transformed without destroying meaning and inhibiting the comprehension of how these sentences relate serially to one another.…”

“…These two researchers found that students reading the de-transformed (no nominalization) sentences inserted more correct nouns, verbs, adjectives, and adverbs on the cloze test blanks than the other students reading the transformed (nominalized) sentences. In another study, Coleman (1964) again used cloze tests to compare 48 college students on their comprehension of prose written in two versions (unsimplified versus simplified by de-transforming four structures: nominalization, relative clause, passive voice, and grammatical deletion). In four different tests, Coleman found that material simplified in this way produced significantly higher cloze test results.…”

Measuring the Readability of Sustainability Reports: A Corpus-Based Analysis Through Standard Formulae and NLP

“…Higher lexical density can lead to higher textual complexity (Halliday 1989, Harrison & Bakker 1998 due to a higher conceptual load. We quantify subordination as the average number of subclause-introducing elements per sentence, which serves as a syntactic complexity measure (Beaman 1984, Dell'Orletta et al 2014, and parse tree depth as the average number of levels in a sentence parse tree, which can indicate complexity and cognitive load (1964, Dell'Orletta et al 2014) Finally, we measure the average number of passive structures per sentence, which the SEC advises against in its Plain English guidelines (Securities and Exchange Commission 1998) A deep-level syntactic analysis is impossible without a parsed corpus and manually parsing a multimillion word corpus is not viable, and infeasible without Natural Language Processing technology capable of automatically analysing text. NLP tools are significantly more technically demanding to implement than readability formulae, but do allow for finer-grained analysis.…”

Die Erforschung der Verständlichkeit von normativen Texten. Eine kritische Darstellung von Modellen, Methoden und Ergebnissen

“…"Nominalst!! 4 * (Coleman 1964), "Satzverschachtelung" (Früh 1980: 179 ff. ), "Satzlänge 44 (Klauer 1984;Evans 1972/73;Wieczerkowski et al 1970;Coleman 1964).…”