Final Report Summary - NUMBER GRAMMAR (Exploring the influence of number cognition on number grammar: a cross-linguistic investigation of Hebrew and English)
Various studies in the number cognition domain claim that enumeration takes place in two distinctive processes: subitising, a fast and accurate process used for numerosities up to three objects (and possibly even four), and counting, a slower and less accurate process for numerosities of four objects and more. It seems likely that the fast and almost flawless cognitive process within the subitising range can support the grammatical number system. The research investigates, in Hebrew and English, the possibility that number language depends on number cognition.
In particular, the goals of the research are to explore:
1. Whether the almost flawless cognitive process within the range of one to three items can support the grammatical number system in a way that differs from the use of the grammatical number system within the range of more than three items.
2. Whether the distinction between the conceptual and grammatical number in the case of mass nouns ('toast') and invariant plurals ('scissors') influences the grammatical number system.
3. To which extent the differences in the morphological and grammatical attributes of Hebrew and English lead to cross-linguistic variations in (1) and (2).
Experiments 1 and 2 aimed to assess the effect of notional number on the process of object enumeration in Hebrew (exp. 1) and English (exp. 2). In exp. 1, two norming surveys were conducted to select the stimuli (the photos and their appropriate naming), from which a list of 246 items were constructed and were used both in exp. 1 and exp. 3. In the actual experiment, participants (N=72) were presented with arrays of 1-6 objects on a computer screen (all objects within an array were identical). They were asked to conduct a numerosity judgment by indicating how many objects were on the display. Responses were made with button presses on buttons labelled with numbers from 1 to 6. The instructions to the participants emphasised both accuracy and speed of response. Response latencies were measured. The conditions included regular count nouns, invariant plural nouns, and mass nouns. The corrected response latencies in the case of correct enumeration demonstrated the expected outcome - a disparity between the subitising (1-3 objects) and counting range (4-6 objects). This finding replicated classic findings (e.g. Trick and Pylyshyn, 1993), but with complex colour photos, and therefore laid the foundations for the naming experiment.
Exp. 2 (N = 72) extended the Hebrew object-enumeration results obtained in exp. 1 to English. One norming survey was conducted to select the stimuli, from which a list of 288 items was constructed and was used both in exp. 2 and exp. 4. 165 of the photos were also used in the Hebrew experiments. The procedure of the experiment was identical to the one described for exp. 1. The results resembled the ones found in exp. 1. Therefore both the Israeli and the American participants exhibited the typical findings in the field of number cognition that distinguish between the apprehension of the amount of 1-3 and 4-6 objects.
Experiments 3 and 4 aimed to assess the effect of notional number on the process of object naming in Hebrew (exp. 3) and English (exp. 4). In exp. 3 (N = 72) participants were presented with the same stimuli presented in exp. 1. For example, presented with a display depicting 3 computers (/max?ev/), participants were asked to name the items (i.e. they are expected to say /max?evim/). As in the first project, the requirement for speed and accuracy was emphasised. Naming latencies were measured.
There were two main results:
(a) We found a decrease in naming latency between numerosities 2-3 and 5-6 for regular count nouns, which was not found in the mass nouns or the invariant plurals.
(b) We found an increase between naming one and more than one object in the mass nouns (more specifically, within the small number range), but not in the regular count or the invariant plural nouns.
We link the difference found between numerosities 2-3 and 5-6 to the mechanisms of number apprehension which were pre-established in exp. 1. However, the fact that the discontinuity was in the form of a decrease in latency between the two ranges suggests that within the large number range, the counting procedure was not automatically triggered.
Exp. 4 (N = 72) served as the English experimental equivalent to exp. 3. The procedure was identical the one described for exp. 3. Overall, naming in English was faster than in Hebrew. Nonetheless, latency characteristics replicated the main properties found in exp. 3.
In particular, the goals of the research are to explore:
1. Whether the almost flawless cognitive process within the range of one to three items can support the grammatical number system in a way that differs from the use of the grammatical number system within the range of more than three items.
2. Whether the distinction between the conceptual and grammatical number in the case of mass nouns ('toast') and invariant plurals ('scissors') influences the grammatical number system.
3. To which extent the differences in the morphological and grammatical attributes of Hebrew and English lead to cross-linguistic variations in (1) and (2).
Experiments 1 and 2 aimed to assess the effect of notional number on the process of object enumeration in Hebrew (exp. 1) and English (exp. 2). In exp. 1, two norming surveys were conducted to select the stimuli (the photos and their appropriate naming), from which a list of 246 items were constructed and were used both in exp. 1 and exp. 3. In the actual experiment, participants (N=72) were presented with arrays of 1-6 objects on a computer screen (all objects within an array were identical). They were asked to conduct a numerosity judgment by indicating how many objects were on the display. Responses were made with button presses on buttons labelled with numbers from 1 to 6. The instructions to the participants emphasised both accuracy and speed of response. Response latencies were measured. The conditions included regular count nouns, invariant plural nouns, and mass nouns. The corrected response latencies in the case of correct enumeration demonstrated the expected outcome - a disparity between the subitising (1-3 objects) and counting range (4-6 objects). This finding replicated classic findings (e.g. Trick and Pylyshyn, 1993), but with complex colour photos, and therefore laid the foundations for the naming experiment.
Exp. 2 (N = 72) extended the Hebrew object-enumeration results obtained in exp. 1 to English. One norming survey was conducted to select the stimuli, from which a list of 288 items was constructed and was used both in exp. 2 and exp. 4. 165 of the photos were also used in the Hebrew experiments. The procedure of the experiment was identical to the one described for exp. 1. The results resembled the ones found in exp. 1. Therefore both the Israeli and the American participants exhibited the typical findings in the field of number cognition that distinguish between the apprehension of the amount of 1-3 and 4-6 objects.
Experiments 3 and 4 aimed to assess the effect of notional number on the process of object naming in Hebrew (exp. 3) and English (exp. 4). In exp. 3 (N = 72) participants were presented with the same stimuli presented in exp. 1. For example, presented with a display depicting 3 computers (/max?ev/), participants were asked to name the items (i.e. they are expected to say /max?evim/). As in the first project, the requirement for speed and accuracy was emphasised. Naming latencies were measured.
There were two main results:
(a) We found a decrease in naming latency between numerosities 2-3 and 5-6 for regular count nouns, which was not found in the mass nouns or the invariant plurals.
(b) We found an increase between naming one and more than one object in the mass nouns (more specifically, within the small number range), but not in the regular count or the invariant plural nouns.
We link the difference found between numerosities 2-3 and 5-6 to the mechanisms of number apprehension which were pre-established in exp. 1. However, the fact that the discontinuity was in the form of a decrease in latency between the two ranges suggests that within the large number range, the counting procedure was not automatically triggered.
Exp. 4 (N = 72) served as the English experimental equivalent to exp. 3. The procedure was identical the one described for exp. 3. Overall, naming in English was faster than in Hebrew. Nonetheless, latency characteristics replicated the main properties found in exp. 3.