Experiment design

SMI Red-x eye tracker was used for eye movement data collection (binocular system, sample rate = 60 Hz, accuracy = 0.4°, head movement 40×20 cm, operating distance = 60–80 cm). We set the minimum fixation duration at 100 msec, in accordance with previous research (Sharmin, Špakov, Räihä, 2012).

Eye movement events data were collected for parts of the text, marked as Areas of Interest (AOIs) in BeGaze 3.0 software. We regarded AOIs with graphological foregrounding, following P. Simpson’s notes on graphological foregrounding (Simpson, 2004), with accentuated features of typeface (bold, italics, caps). Measures included AOI glances count and AOI initial fixation duration.

Before the eye tracking experiment, the participants were subjected to psychological Familiar Figures Test (MFFT) (Kagan et al., 1966). The test is aimed at measuring “reflectivity-impulsivity” cognitive style, also referred to as conceptual tempo (Zhang, Sternberg, 2012). The two test measures are 1) latency (time taken to respond) and 2) accuracy (number of errors). Based on the latency and accuracy score related to the median of the sample, a subject’s result is attributed to either “reflective” (long latency, high accuracy) or “impulsive” group (short latency, low accuracy). The construct “impulsivity-reflectivity”, along with other cognitive styles, is used in the field of learning ability (Ehrman, Leaver, 2003). In visual perception studies eye movement patterns of impulsive and reflective subjects have been distinguished: reflective subjects tend to demonstrate linear patterns of visual search (distinct search pattern, few changes of saccade direction), while impulsive subjects showed non-linear patterns (indistinct search pattern, frequent changes of saccade direction) (Blinnikova, Izmalkova, 2017).

SMI BeGaze 3.0 software was used for raw data processing, and IBM SPSS Statistics 20 was used for data analysis.

Results

We used K-means cluster analysis to define two groups with more “impulsive” and more “reflective” cognitive style, based on two MFFT measures — latency (time taken to respond) and accuracy (number of errors). In line with studies on impulsivity/reflectivity (Kholodnaya et al., 2013) a 4-cluster solution was applied to determine whether all possible variants would be manifested in the sample: long latency/high accuracy, short latency/low accuracy, long latency/low accuracy, short latency/high accuracy. However, only the first two groups were present, which resulted in dividing the sample into two clusters: 9 “impulsive” subjects (fast and erratic answers) and 7 “reflective” subjects (slow and correct answers). The 2-cluster solution is given in details below (see Table 1).

Significant distinctions were found in eye movement characteristics of Impulsives and Reflectives when graphological foregrounding was concerned. We marked AOIs with graphological foregrounding, based on features of typeface (bold, italics, caps), following P. Simpson’s notes on graphological foregrounding (Simpson, 2004). The AOIs for this purpose were selected based on salient features of typeface.

Impulsive subjects tended to visit the highlighted text fragments more often than reflective subjects, as confirmed by Mann–Whitney U test for Glances count (z = -3.95, p < 0.01). Impulsive subjects showed significantly more AOI glances count (Mdn = 1.00), as compared to AOI glances count of reflective subjects (Mdn = 0.00). Impulsive subjects also tended to make longer initial fixations on AOIs (Mdn = 165 ms), than reflective subjects (Mdn = 128 ms), as confirmed by Mann–Whitney U test for AOI first fixation duration on AOIs (z = -2.82, p < 0.01).

This effect was also found in case graphological foregrounding, set with typeface, was accompanied with spatial features of the AOI, if AOIs were situated on a separate line. Impulsive subjects tended to make longer initial fixation duration on the AOIs (Mdn = 178 ms), and visit the AOIs more often (Mdn = 1.00), as compared with reflective subjects (Mdn = 139 ms for fixation duration and Mdn = 0.00 for AOI glances count), as confirmed by Mann–Whitney U test for AOI first fixation duration (z = -2.38, p < 0.05) and Mann–Whitney U test for AOI glances count (z = -3.6, p < 0.01).

Experiment design

The experiment testing modifications in discourse construal within the two participants’ groups, the Impulsives and the Reflectives, was preceded by the elaboration of the system assessing their modifications (of the stimuli) in speech. Since we considered three basic groups of semantic foregrounding parameters, the construal of participants (also objects and actions), event and perspective, the developed parameter system specified these three groups (Wårwik, 2004; Iriskhanova, 2014; Kiose, 2021). We then applied it to contrast the semantic patterns of foregrounding participants, events and perspective in stimuli texts and the participants’ discourse responses to detect the construal differences.

The study involved the following steps.

1. Specification of participant / object / action, event and perspective modifications (and parameter coding).
2. Annotating the participants’ discourse responses (16 participants’ responses were annotated, approx. 32 min.).
3. Contrastive analysis of modification parameter activity and parameter synchronized activity in the discourse responses of the Impulsives and the Reflectives.

The modification parameter system of participant / object / action construal included 9 parameters for participants, for instance, giving full participant’s name (101), participant’s name (102), modified participant’s name (103), 3 parameters for objects, exact object naming (201), modified object naming (202), lack of object mentioning (203), and 4 parameters for action, exact action mentioning (301), generalized action mentioning (302), lack of action mentioning (303), additional characteristics of action (304). The modification system for event construal included 5 subgroups assessing event mentioning on the whole (401–405), event characteristics mentioning (501–506), event reporting by author and participants (601–604), time construal (701–702), space construal (801–802). The system for perspective construal included subjectivation (901), objectivation (902), and intersubjectivation (903).

The total number of modification parameters applied for discourse modification assessment in the participants’ responses was 38.

Since there were 24 Areas of Interest corresponding to the zones of author’s interdiscourse, there were 24 zones of author’s and characters’ interdiscourse or the units of analysis which manifested the change of participants and objects, change of time and location, change of type of relations among participants (Labov, Waletsky, 1967; Labov, 2013). Below, we will show the example of annotating the discourse modifications in 6 participants’ responses.

Stimulus:

Софья Васильевна. Но ведь в шляпе. В шляпе – заметьте! Отец семейства сорока пяти лет!

Sofja Vasiljevna. But in a hat. In a hat – notice! Father of the family of forty-five years old!

Experiment participant:

Другие сочли это смешным нелепым в том смысле, что человеку-то уже сорок…за сорок и это очень странно с его стороны;

Others found it funny ridiculous in a sense that a person is forty…over forty and it’s very odd for him.

We may notice the use of the following modifications: 1) generalization in participant’s construal (104): a person, others; 2) generalization in characteristics (503): a person is forty…over forty; it’s very odd for him. 3) objectivation (802): others found it <…>, a person is <…>, it’s very odd for him.

Experiment participant:

Потом добавляют, что снова, да и заметьте в шляпе отец семейства сорока пяти лет;

Then they add that again and yes notice in a hat head of the family of forty-five years old.

The fragment displays the following modifications: 1) exact characteristics (501): in a hat; 2) objectivation (902): they add; 3) reporting the event in the present (601); 4) indirect event reporting (603): yes notice in a hat head of the family of forty-five years old.

Experiment participant:

Главу семейства там, ему около сорока пяти лет;

Head of the family well… he is around forty-five.

Here the modifications are: 1) modification of characteristics (502): head of the family; around forty-five; 2) objectivation (902); 3) reporting the event in the present (601).

Experiment participant:

Рассказывает за столом об отце семейства.

At the table talks about father of the family.

The modifications include: 1) exact characteristics (501): father of the family; 2) space location (801): at the table; 3) objectivation (902); 4) reporting the event in the present (601).

Experiment participant:

Мужчину сорока пяти лет крайне странного эксцентричного поведения.

A man of forty-five of extremely odd eccentric behaviour.

We may notice here: 1) generalization in participant’s construal (104): a man; 2) exact characteristics (501): forty-five; 3) generalized characteristics (503): extremely eccentric behaviour; 4) objectivation (902).

Experiment participant:

Ему сорок пять лет и для его возраста вообще довольно странное поведение;

He is forty-five and for his age in general this is rather odd behaviour.

The modifications include: 1) exact characteristics (501): forty-five; 2) generalised characteristics (502): for his age in general this is rather odd behaviour; 4) objectivation (902): for his age <…>; 5) reporting the event in the present (601).

The given examples of stimulus modifications manifest common features for the participants, for instance mentioning exact characteristics (father of the family, forty-five years old, in a hat), generalized information about event (others found it funny ridiculous, it’s very odd for him, a man of extremely odd eccentric behavior), and objectivation (he is forty-five and for his age <…>, talks about father of the family, then they add). The differences appear in the use of generalization in participant’s construal (a person, a man), modification of the characteristics (head of the family), not exact characteristics (over forty, around forty-five). Contrastive analysis was then held applying R-Pearson to test the contingency between the relative modification values in two groups. The modifying parameters variance was then tested with One-Sample T-Test.

Results

R-Pearson of relative modification values of the Impulsives’ and the Reflectives’ responses is very high (r (36) = 0.96, p < 0.05), which suffices to claim that in general the modifications are similar.

The analysis of parametric activity has detected that the most active parameters for the impulsive respondents are mentioning of subject’s full name, use of lexeme with wider semantics, stating subject’s role in the play, generalized information about action, subjectivation and objectivation whereas for the reflective respondents the most prominent parameters turned out to be full name of subject, not full name of subject, modified subject’s name, use of lexeme with wider semantics, stating subject’s role in the play, exact action mentioning, generalized information about action. The two groups of respondents – impulsive and reflective – displayed similarity in such parametric activity as lack of subject’s mentioning in the narration, lack of action mentioning, present tense narration. The discourse profiles of relative mean values of modifications of the Impulsives and Reflectives are given in Figure 1.

One Sample T-test has revealed the modifications displaying similarity in both groups (with p < 0.5). Thus, the almost identical data between the two groups of the respondents are found in the following modifications: giving full participant’s name (101) with t (36, 1) = 12.9 at p = 0.049, not mentioning the participant (109) with t (36, 1) = 864.44 at p < 0.001; generalized information of action (302) with t (36, 1) = 25.41 at p = 0.025, not mentioning the action (303) with t (36, 1) = 48.11 at p = 0.013. Several common modifications typical of the Impulsives and Reflectives can be demonstrated in the following example:

Stimulus:

Софья Васильевна. Я обомлела, когда вошла.

Все смеются ещё сильнее.

Sofia Vasilievna. I was dumbfounded when I entered.

Everybody is laughing more gaily.

Experiment participant (impulsive):

Все смеются. Вызывает у них смех.

Everybody is laughing. Causes their laughter.

Experiment participant (reflective):

Все смеются над этой ситуацией. Они обсуждают его странное поведение.

Everybody is laughing at this situation. They are discussing his strange behavior.

Both discourse fragments are similar in manifesting generalized information of action (302), since they both omit its repeated or rather continuous state in еще сильнее.

The following example displays the responses both giving full participant’s name (101) and generalized information of action (302):

Stimulus:

МАТЬ. (задумчиво). Да, пожалела… Его я пожалела. Правда, потом я всю жизнь проклинала себя за тогдашнюю свою жалость.

Пауза.

MOTHER. (thoughtfully). Yes, felt pity…felt pity for him. But afterwards I blamed myself for that pity the rest of my life.

Experiment participant (impulsive):

Мать заканчивает этот диалог говорит, что да раскаивается.

Mother finishes the dialogue, says yes she feels pity.

Experiment participant (reflective):

Мать говорит, что пожалела, да.

Mother says she felt pity, yes.

The variance the participants displayed was not statistically verified, however we may demonstrate some modifications which are more typical of either group. For instance, the differences are observed in the use of generalization in the participant’s construal (104), character specification (107), and generalized characteristics (503), which are both present more frequently in the Reflectives’ speech. These modifications display opposite construal processes, generalization and specification; however, they are manifested towards different event components. We may notice that the Reflectives give more specific characteristics about the participants and events; at the same time they also cope with making conclusions and providing generalizations. As for the Impulsives, they demonstrate more frequent use of lack of event mentioning (405), event construal in the present (601), and objectivation (902). These results suffice to claim that when the Impulsives construe the event they more often do not relate the contingent events following the main event line, make the event more dynamic by recounting it in the present, seldom presenting their personal opinion on the event as a whole.

The following example demonstrates this difference.

Stimulus:

Женский голос. Открой, детка, открой. Это я.

Витя открывает дверь, долго смотрит, затем впускает соседку Анну Степановну.

Анна Степановна (Обращается к Гране.) Девочка-то спит?

Woman’s voice. Open, baby, open. It’s me.

Nina opens the door, looks steadily for some time, then lets the neighbour Anna Stepanovna in.

Anna Stepanovna (addresses Granya). Is the girl sleeping?

Experiment participant (impulsive):

К ним прибегает Анна Степановна.

To them runs to my mind Anna Stepanova.

Experiment participant (impulsive):

Ей открывают. Она Анна Степановна.

The door is opened for her. She is Anna Stepanovna seems to me.

Experiment participant (reflective):

К ним стучится их соседка Анна Степановна э ээ её впускает Нина, э Степановна интересуется, как дела у их ребенка, как я понял.

Their neighbour is knocking the door er er Nina lets her in er Stepanovna asks how is their baby as far as I have understood.

Experiment participant (reflective):

Cоседка, работающая сторожем, пришла в семью Гавриловых, и э стала соответственно обсуждать ребёнка Гавриловых.

The neighbor working as a doorkeeper came to the family of the Gavrolovs, and er started consequently to discuss Gavrilovs’ child.

Experiment participant (reflective):

Здесь происходит встреча. Соседка приходит в гости к Гавриловым, это семья…э… и они, по-моему, разговаривают о детях.

Here there happens a meeting. The neighbour comes to meet the Gavrilovs, it is a family…er… and they it sems to me are talking about children.

The responses of the Impulsives display higher objectivation (902) since they do not involve any indicators of their personal opinion (in the responses of the Reflectives we see the subjectivation как я понял, по-моему). We may also notice that the responses of the Reflectives involve more microevents (это семья, работающая сторожем).