Mental Verbs In Different Domains

The task was presented to two-hundred and seven Italian students, divided in the following school levels:

- Grade 3 (eight years of age) and 5 (ten years of age) of Primary School (respectively: 20 males plus 14 females and 20 males plus 24 females);

- Grade 2 (twelve years of age) of Junior High School (19 males plus 17 females);

- Grade 2 (fourteen years of age) of Senior High School (20 males plus 22 females);

- undergraduates attending different faculties (25 males plus 26 females, ranging in age from twenty to twenty-five years of age).

All students lived in a Lombardia, a region of northern Italy.

The first two subsamples (Grade 3 and 5) were constituted in order to study the development of metarepresentational competencies in children younger than ones considered in Astington & Olson's (1990) study, in which the youngest subsample was drawn from Grade 6 children; we can presume that in earlier ages children should have acquired adequate metarepresentational abilities and basic literacy skills needed to perform in the task. The first four subsamples were separated by 2 years of age in order to evaluate very precisely possible developmental changes. Undergraduates were included as the alleged endpoint in the development of the competence under investigation; their performances should be used to evaluate possible gaps observed in the preceding school levels.

Items presented to half primary school pupils were conjugated in the grammatically correct way (using the subjunctive mood of the verb where needed); items presented to the remaining primary school pupils were conjugated using the same verbal mood appearing in the story. We devised this procedure believing that younger subjects could be affected by the grammatical "surface" of the verb in choosing their answer among the four proposed verbs1.

The researcher explained to students that an anonymous, non-evaluative task will be presented to them. Students were asked to give only one answer for each item. The task was individually administered without temporal limitations.

The whole set of trials was submitted to students over two days: the first day the task concerned one domain of knowledge; the following day the other two domains. The order of the presentation of three domains has been counterbalanced across participants.

1 In Italian the form of the subjunctive mood of verbs is different from the form of the indicative mood.

Score "1" was assigned to each correct answer and score "0" to each wrong answer. For each participant a total score (minimum = 0; maximum = 9) was computed for each domain by counting the number of correct responses he or she answered in that domain. For instance, if a student identified the correct verb in three stories out of the nine provided, the score "3" was assigned to that student.

As far as the possible effects of verbal mood are concerned, we found that neither verbs' conjugation nor interaction between this factor and school level have been shown to be significant.2 We did not obtain significant gender effects except in three cases: to admit in folk psychology and in history and to doubt in mathematics: in all these cases females outperformed males. Analyses of total scores in each domain showed that in folk psychology and mathematics females were significant more correct than males.

Comparing students from humanistic faculties with students from scientific faculties, a significant difference emerges only in the case of the verb to assume in the domain of mathematics, where students from humanistic faculties were significantly more correct than students attending scientific courses. Furthermore, humanities students performed significantly better than scientific students also as regards total scores.

By considering the developmental trend, we notice that the acquisition of metarepresentational verbs is a complex process, lasting till advanced school grades. In fact, even undergraduates did not reach perfect performance (only the verb to admit in folk psychology domain obtained the maximum of correct answers). Undergraduates succeeded within wide ranges of variation (especially for metacognitive verbs), reaching higher levels of performance as far as metalinguistic verbs are concerned (Table 1). The developmental trend varies both within the same verb in different domains and between different verbs in the same domain. Post hoc analyses showed a variety of patterns of homogeneous subgroups of school levels. For instance, significant differences occurred between primary students (Grade 3 and 5) and the older ones (to remember in history); only Grade 3 primary children performed significantly worse than other participants (to hypothesise in history); undergraduates outperformed significantly all the other levels (to predict in mathematics); Primary School pupils (Grade 3 and 5) gave a significant lower number of correct responses than Junior School students who, in turn, gave a lower number of right responses than High School students and undergraduates (to conclude in folk psychology) (Table 1). The overall picture suggests that metalinguistic verbs are mastered before metacognitive verbs and that the former ones are in general easier to identify than the latter ones as proved by the overall mean of correct responses (0.62

2 Detailed statistical analyses are reported in Antonietti, Liverta-Sempio & Marchetti (1998a; 1998b).

vs. 0.50) and by the ranges of the means which is from 0.49 to 0.77 in the first case and from 0.33 to 0.61 in the second case.

Table 1- Mean Numbers of Correct Answers under each School Level for each Verb in each

Domain (First Study)_

Verb Primary Primary Junior High Senior High University

School School School School

Grade 3 Grade 5 Grade 2 Grade 2

To admit psychology 0.61 0.87

history 0.71 0.66

mathematics 0.43 0.57

To assume psychology 0.32 0.49

history 0.68 0.70

mathematics 0.27 0.45

To conclude psychology 0.20 0.28

history 0.27 0.55

mathematics 0.11 0.32

To confirm psychology 0.61 0.57

History 0.47 0.57

mathematics 0.36 0.49

To doubt psychology 0.23 0.47

History 0.26 0.45

mathematics 0.16 0.51

To hypothesise psychology 0.14 0.30

History 0.18 0.45

mathematics 0.05 0.06

To infer psychology 0.07 0.36

History 0.29 0.39

mathematics 0.16 0.51

To predict psychology 0.25 0.64

History 0.36 0.52

mathematics 0.20 0.13


Primary Primary Junior High Senior High University School School School School

Grade 3 Grade 5 Grade 2 Grade 2

To remember psychology History mathematics

Total psychology

History mathematics

If we consider the total scores obtained in each domain (Table 1), we realise that in mathematics the highest number of significant increases among school levels emerged (Grade 3 of Primary School vs. Grade 5 Primary and Junior School vs. High School vs. University); at the opposite, only a significant increase occurred (from Primary School to the subsequent grades) in history. The three domains shared the presence of a significant change at the primary school level: from Grade 3 to 5 in folk psychology and mathematics; from Grade 5 and junior school in history.

As regards the three domains of knowledge, in some verbs the metarepresentational ability under investigation was modulated by the domain. For instance, as far as the verb to hypothesise was concerned, we noticed that the rates of correct identifications were markedly lower in mathematics (0.07) as compared to folk psychology (0.39) and history (0.53) and that such a trend was present at each school level. The same trend can be observed in the verb to assume, even though marked differences between mathematics and the other domains have recorded from junior school to university but not in primary school.

Some verbs are more difficult overall such as to assume, to doubt, and to conclude, but there is considerable variability across items and domains. The wide range of variation in the acquisition of the metarepresentational verbs shows that we dealing with "families" of verbs inherently heterogeneous. The meaning of these verbs appears to be strongly situated and depending on the specific context of occurrence. We suggest that it can depend on the higher "concreteness" of thinking about of speech acts compared with thinking about thinking: the former are more instantiated in social relationships and exchanges, whereas the latter, though originating from communication, ends in internalisation and abstraction.

Table 2 - Mean Numbers of Correct Answers to each Verb Ordered in Term of Difficulty According to the Total Scores (Second Study)_




Metacognitive verbs

to assume




to remember




to doubt




to infer




to hypothesize




Metalinguistic verbs

to confirm




to admit




to conclude




to predict




0 0

Post a comment