Linguistics 897 TA Training/Practicum Course—L’École de SpecGram SpecGram Vol CLXXX, No 1 Contents Teaching Linguistics in High School—Sven Slater and Ollie Bickford

A Classroom-Based Study of Semantic Entailments:
Definiteness, Specificity, and Anaphoric Reference in English, French, and German

by Skiffy Bafflegab
Lecturer in Semantic Theory,
University of the Four Corners, New Mexico Campus


Definiteness and specificity are longstanding topics of research in semantics; in a pre-theoretical view, definiteness indicates that the speaker asserts the hearer can identify the referent of a noun, while specificity indicates only that the speaker herself can identify it. For the most part, specificity is important with indefinite nouns: ‘I’m looking for a French book. Do you have it (spec)/one (non-spec)?’ From a theoretical perspective, specificity has been studied in regard to such matters as scope over operators, referentiality, and partitivity.

Literature Review

While there has been extensive research during the past two decades on the role of specificity and definiteness in disambiguating anaphoric reference in numerous languages within certain restricted contexts, most of it is methodologically fatally flawed, theoretically null, and practically useless (Bafflegab 2016). Indeed, the only studies worth scholarly attention are currently unpublished (Bafflegab 2013, 2014, 2015a, 2015b, 2017a, 2017b).


This study was conducted during the 2016 fall semester in the Introductory Semantics course at the researcher’s university. Participants consisted of 45 sophomores majoring in linguistics (17 male, 28 female), with a mean age of 19 years 10 months (σ = 3.7 months). The study consisted of a pre-course questionnaire and three experiments on phenomena involving English, French, and German anaphoric reference, respectively. The pre-course questionnaire was used to determine the starting language capacities and interests of the students (Tables 1 and 2).

Table 1. Descriptive statistics of the participants: Language ability

Table 2. Descriptive statistics of the participants: Interests

Participants were allowed to drop out of the experiment if they wished; however, as the course is a prerequisite in the department, none did so. The experiments concerned the judgment of 184, 162, and 157 sentences involving the interaction of definiteness or specificity and anaphora in English, French, and German, respectively, collected from the major papers and books on the subject, both by the researcher and the misguided; for each sentence, four options were given for reference: a, b, both, and ambiguous. As preparation for the experiments, students were assigned the most important papers on the subject, which were discussed and for the most part eviscerated in class. The results of the experiments were determined by how far from the predictions made by the SDAT (Specificity, Definiteness, and Anaphoric Tracking) Theory (Bafflegab 2012) the answers for each student fell by the procedure of counting up the correct answers. Incomplete tests were not excluded, as they provided a useful measure of participants’ cognitive abilities that forms a major part of SDAT Theory.


The experiments spectacularly confirmed the predictions of this theory. On the first experiment (“First Examination”), the results showed a distribution very close to normal with a mean of 47.3 correct answers (σ = 12.1). On the second experiment (“Examen de Mi-Trimestre”), the results showed a fair degree of skewness (μ = 16.7, σ = 3.1, γ1 = −2.1). On the third experiment (“Abschlussprüfung”), the results were hard to fit to a distinctive curve due to the discreteness of the scale (μ = 3.1, σ = 1.7, γ1 = −0.6). The results bear out the predictions of Applied Second-Language SDAT by showing that in any language, the results are always worse than predicted by random chance. The grade for each student was determined in usual academic fashion by taking the grades defining the quintiles of the results on the first experiment as the cutoff for the grades on all experiments and grading the students accordingly; the experiments were weighted as counting for 20%, 30%, and 50% of the final grade, respectively. As expected, all students failed the course spectacularly with a distribution very similar to the hypothesized results, confirming the predictions made by the researcher almost perfectly (R2 = 0.984).


1 As SpecGram does not allow citations of unpublished papers, this section is left blank. —Eds.

Linguistics 897 TA Training/Practicum CourseL’École de SpecGram
Teaching Linguistics in High SchoolSven Slater and Ollie Bickford
SpecGram Vol CLXXX, No 1 Contents