Keywords

1 Introduction

Mechanisms of understanding of a text, which are the basis of text processing while reading, are among the main problems of modern psycholinguistics. Large number of experiments has shown that the understanding of the text (written or oral) is a complicated and multi-step process. Thus, understanding of a written text includes a plurality of additional variables, such as, recognition of letters, drawings and structural components. In modern linguistics, the question of an influence of a text type on text perception and comprehension has been studied extensively [16]. M-A.Laane identifies the following variables affecting the “readability” of the text: content (abstracts, organization, coordination), style (semantic and syntactic elements), the format (design, layout, illustrations), organization (paragraphs, structure, headings) [3]. William H. DuBay notes that the text style can be experimentally verified. According to his opinion in the study of different texts there are following indicators of readability: the average length of sentences (in words), the percentage of “simple” words, the number of different “complicated” words, the number of pronouns (1, 2 and 3 person), the minimum length of a sentence (in syllables), the maximum length of a sentence (in syllables), the average sentence length (in syllables), the percentage of monosyllabic words, the percentage of polysyllabic words [4].

Mortimer Adler in his work “How to Read Books” made a very important conclusion, saying that reading is an active process [7]. At first glance, you may divide all speech activity on two processes—active and passive. In such opposition it is quite obvious, that speech production (writing and speaking) is an active process, and reception (reading and listening) is a passive process. But it is impossible to assert that while processing a text (written or oral) the communicant is passive and doesn’t act. Communicant makes a number of actions while reading or listening a text: receives information, decodes it, interprets and comprehends it in a particular way. Reading it is not an easy task, as we naively believe. It consists of different stages, on each of these stages you can improve and hone your skills [7]. Thus, reading it is an active verbal and cogitative process consisting of several stages. The text is a complex and multifaceted object of perception, “it can be more or less understood holistically, from a minimum of assimilation to the full disclosure of the author’s intention.” The degree of integrity of text comprehension, respectively, depends on the degree of activity and involvement of a reader in the reading process. However, the type of a text will also affect the level of reading comprehension. There are two types of reading: “for information” and “for understanding” as M. Adler suggests. While reading for information we look for the facts in the text, get some knowledge and accept it. While reading for understanding we look for the meanings, concepts and implications and try to combine them with our own worldview. Reading for information is a necessary stage of studying at the starter levels, when demonstration of language units and their functioning is a subject of training and is a main goal of using of the text. However, when, skills are developed and the language material is acquired, the teacher should pay more attention to motivation of students to understand the text, but not just to extract the information from it. Thus, differentiating process of reading on two types, we can also speak about two types of the texts: that don’t demand long processing and efforts for understanding and which main content lies on a surface, and also that which are acquired not at once and demand more time for perception and comprehension [7].

2 Research

2.1 Hypothesis

Detection the features of processing and understanding of texts of different functional styles in reading process was the purpose of the experiment described in this article. The hypothesis is based on the assumption that the processes of perception, including reading, processing and reproduction, in texts of different functional styles are different.

2.2 Material

For a pilot study three functional styles were chosen: scientific, publicistic and colloquial, as representing the greatest interest as a material at the lessons of Russian language and Russian for foreigners. Thus, in order that ease or complexity of the text wasn’t caused by thematic features, and also to output some average result on functional style, but not on concrete unit, 9 texts on three different topics were taken, and each topic was presented by three texts of different functional styles. The choice of a subject was caused by existence of texts of these topics in all selected functional styles, and also by a potential possibility of using these texts at lessons of Russian and Russian as a foreign language. 3 topics connected with daily human life were chosen: thunder-storm, relations of children and parents, cars. Thus, the material of the research contained 9 texts:

  • 3 texts about a thunder-storm (scientific, publicistic and colloquial);

  • 3 texts about the relations between parents and children (scientific, publicistic and colloquial);

  • 3 texts about cars (scientific, publicistic and colloquial).

The functional styles of these texts were defined by experts (Table 1). Seven respondents (4 women, 3 men)—teachers of Russian language and literature—were asked to define to which of the functional styles (scientific/publicistic/colloquial) this or that text belonged.

Table 1 Results of an expert evaluation of a functional style of the texts

There was a unanimous expert decision in determining a functional style for six of nine texts. The greatest interest is the results in an assessment of text 6. Two experts defined this text as belonging to a publicistic style, four of seven—to scientific, one of experts dropped a hint of doubt, choosing between these two styles, but as a result stopped on the publicistic. This text represents an article from the encyclopedic dictionary, respectively, if to judge on a resource, the text is scientific.

For the further analysis of the texts Text Internet project was used [http://ru.readability.io]. This allowed us to calculate a number of indicators of the texts (are presented in Table 2, from the line “number of signs”), and also on their basis to calculate indexes of readability for each text (on the formulas adapted for Russian).

Table 2 The analysis of texts on the category of “readability”

It is possible to notice that indicators in each group (scientific, publicistic and colloquial) are different. The texts considerably differ from each other according to quantitative characteristics.

The analysis of readability parameters of the texts revealed that scientific texts are allocated against texts of publicistic and colloquial style from the formal point of view. Their peculiarities are the following:

  • contain bigger number of the “compound words” demanding bigger amount of time for text processing;

  • contain less number of sentences at an equal number of words in the text, that indicates the complexity of syntax used in scientific texts;

  • are focused on more adult (from the point of view of education) audience—students, whereas texts of two other styles can be apprehended by children.

However, it should be noted that these indicators of “readability” are especially formal and consider only quantitative characteristics, ignoring the qualitative.

The main objective of the research was a checkout if the functional text style is among the readability categories and if it influences the effect of reading perspective. This technique gives an opportunity to see the processes of thinking and language behavior of the person in their interaction and is the most exact for research of reading processes as it allows not only to identify difficulties when reading, but also to find when and where specifically these difficulties arise and are resolved by the reader [8]. Today it is one of the main methods of the simultaneous analysis of a large number of dependent variables for language handling research.

2.3 Subjects

30 native speakers of Russian aged from 17 to 30 years participated in eye-tracking experiment. All subjects had the higher or incomplete higher education.

2.4 Procedure

After the calibration procedure each subject received the following instruction: “Hello, dear participant! Thanks for readiness to take part in experiment. Please, attentively study the text for further retelling”. Then the record of experiment began. Participants of experiment weren’t limited in time and could study each text as much as they wanted for adequate perception of the text with its further reproduction. Eye movements were monitored with an SR Eyelink 1000 plus eye tracker (SR Research Ltd., Kanata, Ontario, Canada) sampling at 250 Hz. The system was configured in ‘desktop mode’ and equipped with a chin rest. While subjects read binocularly, only one eye (the right eye by default) was tracked. Viewing distance was approximately 60 cm. Approximately 2.4 characters were encompassed by 1° of visual angle. We used a 22’ LCD monitor (HP Compaq LA2205 wg) with a refresh rate of 60 Hz (screen resolution 1680 × 1050). When the subject reported about end of this stage of experiment, he (or she) was asked to retell the text. All the retellings were recorded on a dictophone. After each retelling the subject was asked to answer 7 questions checking the degree of text comprehension. Each subject read three texts, according to the number of texts in the protocol (Appendix 2). Each protocol contained three different texts of three different styles. So, if the subject read a text “Car” written in scientific style, on the second step the same subject read text “Children” written in colloquial style, and the third step—the text “Storm” in publicistic style. The texts were randomised among the subjects. We used drift correction each time next text was presented. And we recalibrated in case we saw that it was necessary. It took about 20 min for each subject to participate in the experiment.

3 Results and Discussion

As a result of experiment 90 records of eye movements during reading the texts were made. The averaged values for each text type are in Appendix 3.

In the analysis of the results the following indicators in each text were considered: an average duration of fixations, an average amplitude of saccades, a number of fixations, a quantity of saccades, a reading time, a percentage of regressions (returns to the earlier read fragment) to the total number of saccades.

For statistical data processing Mann-Whitney’s U-criterion was used from SPSS Statistics. Statistically significant difference between indicators of reading of the text was revealed for texts “Storm” and “Children”, namely: reading time, number of saccades and fixations was significantly higher in the texts of scientific style than in texts of colloquial and publicistic styles; and also in reading scientific text about the Storm the percentage of regressions was higher than in colloquial and publicistic texts (See Tables 3, 4 and 5). Bold print in the tables indicates statistically significant differences. We connect the lack of a significant difference in indicators of reading the text about cars with features of Text 6 which is intermediate (see an expert assessment of the text style).

Table 3 Results of comparison of eye-movement indicators while reading the texts of different functional styles on the topic “Car” (p-level)
Table 4 Results of comparison of eye-movement indicators while reading the texts of different functional styles on the topic “Storm” (p-level)
Table 5 Results of comparison of eye-movement indicators while reading the texts of different functional styles on the topic “Children” (p-level)

The analysis revealed that the process of reading the texts of scientific style opposes the process of reading the publicistic and colloquial texts. There were significant differences in the processing of publicistic and colloquial texts.

The analysis of answers to the questions after each text allowed to conclude that subjects are capable to apprehend the main ideas of the texts of different functional styles, but the degree of understanding is different.

Texts of scientific style were more difficult for comprehension. There were significantly more mistakes in the answers after scientific text reading in comparison with the answers after publicistic or colloquial texts. Texts of scientific style demand more time for comprehension and we suppose that degree of their comprehension directly depends on the volume of working memory of a subject. Whereas texts of publicistic and colloquial style stimulate the reader to create his own text and to identify himself with the author of the text [9].

The mistakes made in answers to the questions on comprehension of such texts are connected first of all with a replacement of the read information by an actual information from the reader’s background (Table 6).

Table 6 Percentage of correct answers after the text questions

The analysis of retellings confirms the conclusions drawn above (Table 7).

Table 7 Percentage of retellings that contain factual mistakes in comparison with the original text

Texts of colloquial and publicistic style are reproduced easier: subjects make less mistakes at a statement of the main content of a text, they generalize the received information a little, draw conclusions on its basis and transform, “personalize” the text.

Some retellings contained subject’s comments such as: “as I understand”, “to my mind”, “I think so”, “It seems to me” etc. (Table 8).

Table 8 Percentage of retellings containing subject’s comments

4 Conclusion

Generalizing all received results, it is possible to conclude that strategies of reading texts of different functional styles are different.

We suggest that the following eye-tracking data can be informative when we need to evaluate text readability: amplitude of saccades, number of regressions, fixation duration while searching for an answer in the text. The readability level of the text influences the effect of reading perspective. The results of checking the texts by readability formulas (Table 2) are correlated with eye tracking data and retellings.

The results indicate that the text on the same topic is easier read if it is written in a publicistic style than in a scientific style. There were no significant differences in eye-tracking data between texts written in publicistic style and colloquial style. But the publicistic text is read, processed and reproduced better, than scientific.

These conclusions are important from the point of view of methodology of teaching languages (at least Russian). Scientific text is more difficult for reading and understanding. So, if the teacher selects it for studying in a classroom more pre-text work will be needed. It’s not recommended to take scientific texts for demonstrating grammar or syntactic constructions as it will require much time for understanding the meaning of the phrase or the text, while in colloquial and publicistic texts it’s easier to concentrate on studying grammar.

There are the following perspectives for further research:

  • how do foreign students learning Russian language read texts of different functional styles, are there the same trends as in reading texts by native speakers;

  • how does the instruction affect the reading and comprehension of a text;

  • is there any difference in reading texts with pre-text work and without it;

  • what types of pre-text work are better for text reading and comprehension.