Gypsy Classes and "Mixed Classes" - In View Of the Facts

By Zita Réger
Source: Valóság, 1978/8.

Is it correct for them to exist? Do we need Gypsy classes and if so, what should they be like? - Heated debates on this topic are recurring again and again (the latest of these in the weekly Élet és Irodalom (Life and Literature) in the autumn of 1976 ). Undoubtedly, the relatively new form of educating Gypsy children, which appeared approximately 15 years ago, i.e. the establishment of Gypsy classes and Gypsy schools, was institutionalized in answer to real pedagogical difficulties which had largely determined the school careers of these children having to struggle with socio-cultural, hygienic, psychological and sometimes linguistic problems over the previous decades. The acuteness of these problems has been highlighted by the slightly if at all improving statistics over the years: the failure of a large number of primary school children, and a high proportion of pupils dropping out from school every year. (The statistics of the school year 1974/75 exemplifies this: while 15.1% of all primary school pupils were 1st grade pupils and 11.8% were 8th grade pupils, 22.7% of all Gypsy primary school pupils were 1st grade and only 5.3% were 8th grade pupils. In the same school year, 41.1% of the 1st grade Gypsy children failed to comply with the minimum requirements at school. As compared with the school year 1970/71, little progress had been made.)

The objectives and characteristics of the separate Gypsy study groups and afternoon study groups to be established are described in the Government Decree Number 19, 12598/1962. "The separate Gypsy study groups and whole-day classes are established on a temporary basis. The objective of their establishment is to make it possible for the pupils to continue their studies successfully in normal classes after one or two years (Point 3)". It also states that "a student whose attitude towards learning shows such progress by the end of the school year that enables him or her to achieve in the normal classes and whole-day schools should be directed to continue his or her studies in those classes. (Point 4)"

Those involved in the debates have rather varied opinions whether this organizational form, the "Gypsy class", established on a markedly temporary basis in order to remove disadvantages, will ever achieve its aims and does really serve the purpose that it had originally been established for.

Some argue that a six-year-old Gypsy child, whose original environment is rather non-stimulating and whose command of Hungarian is also insufficient, requires education that is based on perspectives and methods that are different from those of the education of average six-year-old non-Gypsy children, and this can only happen in a separate class for Gypsy children. Children achieve far better results and a significantly lower proportion of children drop out from school if they attend Gypsy schools or classes than if educated in classes where Gypsy and non-Gypsy pupils study together. The establishment of separate study groups - as a special organizational form - is advantageous for Gypsy children's development for other reasons as well: instead of the cumulative failures that they would experience in the circumstances of "mixed" education, here they have a real sense of achievement and do not have to suffer from prejudices against them, either. (I also shared this opinion myself until a couple of years ago. ) Therefore, according to these arguments, it is necessary to maintain Gypsy classes and schools and it is beneficial to establish more.

On the other hand, despite the fact that theoretically, their establishment is based on a correct idea, practically, Gypsy classes are a dead-end-street. Even in villages where non-Gypsy children studying in different grades attend separate classes, the Gypsy classes are all established for children of different grades with different curricula. Therefore the children are disadvantaged as far as their learning conditions are concerned from the very beginning of their school careers. Due to their isolation, the assessment of their achievements is inaccurate and their results cannot be compared to those of pupils studying in normal classes. After years of isolation, it is even more difficult for the children to integrate into "mixed" classes. Schools tend to be reluctant to receive the often over-age pupils with inadequate knowledge and the isolation of Gypsy children is typically prolonged. Thus, in practice, Gypsy classes, originally established to remove the children's disadvantages, often become "primary schools for Gypsy children", with all the disadvantageous consequences concerning the children's education and social position. And lastly, it is rather frequent that Gypsy classes cannot fight the problems of children often having to repeat classes, becoming over-age pupils and dropping out from school in the long run .

In this debate it is not easy to decide who is right because we can avail of very scarce reliable data. Both those in favor of and those rejecting the idea of maintaining Gypsy classes notice one crucial defect: the lack of carefully designed and monitored pedagogical experiments of several years' length that could reassuringly clarify the issue of the efficiency of Gypsy classes.

As it has become apparent in the course of the debates, the only point of reference concerning the efficiency of these classes is still the success or failure measured by analyzing the school marks that the children receive.

However, educators endorse the public opinion that is based on facts and daily experience that the marks that pupils are given do not constitute an objective system of assessment. Different standards and achievements may be behind the same marks in different schools .

In order to assess the efficiency of the different methods, it would be absolutely necessary to make measurements based on objective methods, i.e. numerical comparisons of the achievements of different groups of children whose knowledge, abilities and level of education in the general sense were the same at the beginning of the educational process but have studied in different organizational forms, i.e. in Gypsy and "mixed" classes. Such measurements have not been carried out as yet. As for the assessment of the efficiency of the approximately 200 Gypsy study groups run nation-wide, we cannot even rely on data from comprehensive surveys on their work, educational circumstances or true achievements, either.

The present case study, comparing certain aspects of the achievements of Gypsy children of homogeneous socio-cultural and linguistic background learning in different organizational forms (in Gypsy and in "mixed" classes), wishes to contribute to the settlement of debate described above and also emphasize that it is necessary to carry out a nation-wide survey of the pedagogical efficiency of Gypsy classes. (I will only present an outline of my study here. A detailed analysis of the data collected will be published in my monograph on bilingual Gypsy children.

THE CHILDREN EXAMINED

Forty Gypsy children (26 boys and 14 girls) learning in the Kossuth Lajos Primary School in Esztergom (a town 50 kilometers from Budapest) comprised the first group of children in the survey. In the school year 1976/77, these 1st-7th grade schoolchildren attended two Gypsy study groups taught by one teacher each. (The pupils of the 1st and 2nd grades comprised one group and the pupils of the 3rd, 4th, 5th, 6th and 7th grades comprised the other. The two afternoon study groups were attended by the same two groups of pupils.) Approximately two-thirds of the children examined came from an isolated Gypsy environment. They live in the 9 semi-detached houses in Ságvári Estate and in the "CS Houses", not far from the estate, built closely next to each other. The families of approximately one-third of them - 15 children - live among "Hungarian" families.

20 Gypsy children (10 boys and 10 girls) learning in the 1st-7th grades in the Bécsi úti Primary School in Óbuda (three 3rd district of Budapest) comprised the second group. In this primary school the gradual eradication of the special study groups for Gypsy pupils of different grades started in this school with the guidance of the Budapest Pedagogical Institute and the financial support of the Municipality of Budapest in the school year 1974/75. As a first step, they tried to improve the circumstances and raise the educational standards of the existing study groups substantially by different measures (e.g. by establishing classes with a whole day study programme, providing better material circumstances, recruiting new teachers and providing better hygienic facilities). Then the children were placed into the different normal classes of different grades. A separate afternoon study group for 1st and 2nd grade Gypsy pupils was operated for a year after the introduction of the reforms while the Gypsy pupils of higher grades attended "mixed" afternoon study groups from the beginning.

Among the Óbuda children, there are also some coming from a more isolated Gypsy environment and some from a more "mixed" background, the proportions being roughly the same as in the case of the Esztergom children. 12 out of the 20 pupils live in a small settlement built near the gas works and the railway that is mostly inhabited by Gypsy families. 8 live in different buildings built in not specifically Gypsy areas and in the barracks-like blocks of the Csillaghegy brick factory.

The large majority of both the Esztergom and Óbuda children participating in the survey were over-age. The mother tongue of all of the children was a Gypsy language and they could speak no or very little Hungarian when they first went to school.

I intended to work with as high a number of children among those more or less regularly attending school in the school year 1976/77, the year of the survey, as possible. Comparing the socio-cultural situation of the Esztergom and Óbuda children, I concentrated on two aspects that - according to a recent pre-school maturity test carried out among Gypsy children - profoundly influence children's development (and whose correlation with the children's chances to succeed at school are well-known). These are the parents' level of schooling and the families' housing situation.

In these two aspects no significant differences can be observed between the Esztergom and Óbuda children's circumstances. None of the parents of the pupils participating in the survey accomplished their primary education. In Esztergom, the ratio of illiterate parents is "only" 42% while 81% of the Óbuda parents are illiterate. 42% of the Esztergom parents finished at least one of the 5th-8th grades, which constitute "upper primary" education. None of the Óbuda parents finished any of these grades. On the other hand, as far as their housing is concerned - although the housing conditions of the families of both groups of children are characterised by a relatively high number of people sharing the rooms - the situation of the Óbuda families is somewhat more advantageous: 45% of the Óbuda children and 65% of the Esztergom children live in housing conditions with over 7 people/room.

Coincidentally, and also luckily for the survey, a certain circumstance further increases the similarities between the socio-cultural backgrounds of the two groups of children. Namely, the Óbuda and Esztergom children have very close family relationships: due to extensive endogamy, the majority of the children in Óbuda and Esztergom in the survey have cousins and second cousins also participating in the survey. The reason for this is that the families of the Óbuda children - with only a few exceptions - migrated from Esztergom to the capital city (actually, several of them from the semi-detached houses of the Ságvári Estate) not so long ago. 15 of the 20 children examined in Óbuda were born in Esztergom. Those people who remained in Esztergom and those who migrated to Budapest still maintain close relationships and the families frequently visit each other.

Thus, assumedly, apart from the similarities in the socio-cultural and linguistic backgrounds of the children compared in the survey, there were great similarities in the customs and lifestyles of their families and the whole process of their socialization. Consequently, the differences between the achievements of the two groups of children must have occurred mainly due to the different circumstances at school and the different standards and forms of education.

I used three tests to assess three aspects of the children's level of maturity and their skills. The test of vocabulary applied in the test, adapted by myself following foreign examples, served to measure the children's Hungarian and Gypsy language proficiency. (Although this test was carried out among children whose mother tongue is a Gypsy language, we must emphasize that the mother tongue of the majority of the Gypsies in Hungary is not a Gypsy language but Hungarian . Many Gypsy classes are in fact attended by children whose mother tongue is Hungarian.) There are various reasons for our choice to carry out the survey among these children. Firstly, the official documents concerning Gypsy classes (and the public opinion) largely emphasize the linguistic aspect. Secondly, among Gypsy children of similar socio-cultural background, the integration into the school community is undoubtedly the most difficult for those whose mother tongue is not Hungarian. Lastly, methodologically, it is useful to examine children whose mother tongue is a Gypsy language because it makes it easier to separate the knowledge acquired at home (the language of which is a Gypsy language) and the knowledge acquired at school (and for this reason - in certain cases - they can only relate it in Hungarian). The second test utilized in the survey was a test frequently used by psychologists in Hungary, the so-called Bender Test, which provided a picture of the degree of children's skills of perception of forms and concerning the reproduction of these, i.e. their "visual-motor maturity". Thirdly, the test of their reading skills directly assessed their school achievements.

The tests were administered from May to June in 1977, partly at the schools and partly in the children's homes. Thus, the data collected in the survey and the description of the teaching/educational environment describe the state of affairs in the school year 1976/77.

THE TEST OF VOCABULARY AND ITS RESULTS

I compiled the test of vocabulary following the example of a test administered by the American psycho-linguist S. Ervin-Tripp examining bilingual American adults whose languages were Japanese and English and Italian and English .

In the course of the test children had to name 100 simple objects and things. These were mostly illustrated in pictures (the majority of these pictures can be found in the A-Z children's encyclopedia called Ablak-Zsiráf and the reader of the first grade pupils) but a small set of small-sized objects was also used (a nail, a needle, a mirror, etc.). The 100 test items were glued onto six sheets of paper. They were arranged in a special way, i.e. the items contextually related were preferably placed next to each other (e.g. the kinds of fruit, different animals, the needle and the thread, the hen and the egg). By such classification of the test items, a part of the objects, things, plants and animals to be named were placed in an "environment" and thus gained a context.

The children participating in the test were instructed to name the different objects, things, plants and animals in two languages, i.e. first they named the first 50 items in Hungarian and the second 50 items in their Gypsy mother tongue, and then the first 50 items in their Gypsy mother tongue and the second 50 items in Hungarian. I measured the time that these 4 stages required with a stop-watch. Children who were unable to give the name of a certain object, thing, plant or animal in approx. 15 seconds were asked to describe what it served for or where they saw it (naturally, they were to answer these questions in the language being tested).

The whole of the vocabulary testing was recorded on tapes. The results of the test administered in the method described above answer the following two questions:

1. Which language is the children's so-called dominant language, i.e. the one that they speak better / more fluently? This can be determined by comparing the amount of time necessary for the children to name the 100 test items in the two languages. Obviously, it takes bilingual children or adults shorter to name the items in the language that they speak better and are more proficient in. So, the test primarily measures the "fluency" of the process of naming the items and the relative speed of activating the vocabulary items. Analyzing its results, it is possible to draw conclusions concerning the "linguistic dominance" relations and the bilingual child's or adult's relative language proficiency.

2. Do children know - and can they name in both languages - the different objects, things, plants and animals in the test? From the frequency of errors of naming the items in Hungarian we can draw direct conclusions concerning the linguistic difficulties of the educational process, since the majority of the 100 items turn up - some of them quite frequently - in the 1st grade reader used in Hungarian primary schools.

At the same time, the test data is informative concerning the efficiency of the educational process, for the children are supposed to have learnt about the majority of the items to be named, e.g. types of fruit, animals or parts of the body at the natural science lessons in the 1st-4th grades. As for the data concerning the children's Gypsy language proficiency, it throws light on the material circumstances and linguistic and cultural influences that have shaped the mental development of the children tested and were formative concerning their mental and linguistic maturity.

I administered the test of vocabulary in a control group of children whose mother tongue is Hungarian. 10 Óbuda children of working class families attending the 1st grade of primary school and 20 kindergarten children of 3 to 6 years of age (4 groups of 5 children of the same age in each group) comprised the control group. The 1st grade schoolchildren attended a primary school in Bécsi út, the kindergarten children attended the kindergarten in Munkácsy Mihály utca, a street located in the 6th district of Budapest.

RESULTS

(The ratios of linguistic dominance)

The ratio of linguistic dominance is the quotient of the lengths of time necessary for the children to name the test item in Hungarian and in their mother tongue. I calculated it by dividing the length of the time necessary for the children to name the test items in Hungarian measured in seconds by the length of time necessary for them to name the same items in their mother tongue. Consequently, a quotient over 1 refers to a dominant proficiency of their mother tongue, while a quotient below 1 refers to the dominant proficiency of the Hungarian language. Thus, a quotient of 2 means that naming the 100 test items in Hungarian required twice as much time as the same process in their Gypsy mother tongue. A quotient of or around 1 means that the child's proficiency of the two languages is approximately the same. In psycho-linguistics, this is called balanced bilingualism .

The test results of vocabulary dominance show that the average ratio of linguistic dominance of the children attending parallel grades belonging to the two groups largely differs. The average ratio of linguistic dominance of the group of the Esztergom 1st grade pupils is 1.96. The ratio approximates 2 in the groups of the pupils of higher grades. This indicates that the relative level of the children's Hungarian language proficiency only slightly improves during the years and their Gypsy mother tongue is still their dominant language when they attend the higher grades of the primary school.

The results in the Óbuda groups of 1st and 2nd grade children (where the average ratios of linguistic dominance are 1.25 and 1.32 respectively) already indicate more balanced levels of proficiency. The ratios of linguistic dominance of the groups of 3rd-4th and 5th-7th grades pupils are below 1 (0.88 and 0.89, respectively). (However, the ratio of linguistic dominance of 2.06 of seven-year-old Mária S., the only non-over-age Óbuda 1st grade pupil, approximates the average ratio of linguistic dominance of the Esztergom children!) The data shows that the Hungarian language proficiency of the Óbuda children in the 1st and 2nd grades (i.e. after two or three years of schooling) usually approximates, and that of the children in higher grades reaches and in certain cases even surpasses their proficiency in their Gypsy mother tongue - at least in the respects tested. (The ratio of linguistic dominance slightly below 1 does not necessarily indicate that the children's command of their Gypsy mother tongue is poorer that their command of Hungarian. Their having poorer results in the Gypsy language may have had certain socio-linguistic reasons, e.g. the aspect that the test was administered in the school, where they usually speak in Hungarian or the fact the administrator of the test was a person belonging to the Hungarian-speaking community.)

It is worth comparing the average period of time necessary for the members of the groups of Gypsy children attending parallel classes in the different schools and for the 1st grade and kindergarten children in the control group to name the 100 test item in Hungarian. The average period of time necessary for the pupils in the 1st and 2nd grades in the Óbuda school to name the test items in Hungarian was the two-thirds of the time necessary for the pupils of the Esztergom school attending the same classes. The average period of time necessary for the pupils in the higher grades in the Óbuda school to name the test items was less than half of the time necessary for the pupils in the Esztergom school attending the same classes to do the same. (However, the result of 838 seconds of the already mentioned seven-year-old pupil, Mária S., is very near the average result of the Esztergom 1st grade pupils.) It is worth noting that in this respect, even the average result of the Óbuda 2nd grade pupils (352 seconds) largely surpassed that of the pupils of the 5th-7th grade Esztergom pupils (414 seconds). On the other hand, the control group data shows that even the Óbuda Gypsy children had serious disadvantages at the beginning of their schooling if we compare their results with those of the Hungarian children.

(Details concerning the naming of the test items. Problems of perceptual and notional nature.) Another aspect of the differences of the Hungarian language proficiency of the Óbuda and Esztergom children originating in the different educational circumstances is highlighted by the different proportions of correct answers concerning the names of the different objects, things, plants and animals listed among the test items. (The children's answers concerning the names of the test items were considered incorrect if they did not know the name of the different objects, things, plants and animals or if they named them in the wrong language or using the wrong word.)

In the course of analyzing the results, another aspect that I took into consideration apart from the average number of incorrect answers in the two languages was the number of the items that the children belonging to the parallel classes of the different schools were unable to name in either of the two languages. Examples for this "absolute error ratio" are the answers of an Esztergom schoolgirl who called the goose in the picture csibe (chick) in Hungarian and khajnyi (hen) in her mother tongue.

The difference between the two groups' achievements proved rather significant. The rate of errors of the Hungarian naming process in the case of the groups attending different classes in the Esztergom school was two, three, or in the case of the pupils of the 3rd and 4th grades, even 10 times as high as that of the groups of children attending parallel classes at the Óbuda school. (Again, the results of Óbuda schoolgirl Mária S., who had only been attending school for 1 year at the time of the testing concerning her Hungarian language achievement (43%) and her absolute error ratio of 11% approximate the respective results of 49.7% and 11.7% of the Esztergom 1st grade pupils!)

If we compare the Hungarian language achievement of the Esztergom group with that of the kindergarten control group, we have to acknowledge that, although it is almost unbelievable, the average rate of errors of the Esztergom 1st and 2nd grade schoolchildren (49.7% and 37%) is higher than that of the three-year-old kindergarten children whose mother tongue is Hungarian (32.2%) and the results of the pupils of the 3rd-4th and 5th-7th grades (19.2% and 13.2%) are lower than those of the four- year-old kindergarten children (12.8%).

If we compare the Hungarian language achievement of the 1st and 2nd grade pupils of the Óbuda group (their rates of errors are 21.3% and 11.3%) with that of the control group, we find that their backwardness is also rather significant. However, this disadvantage seems to diminish considerably in the higher grades.

(…)

It is remarkable that the types of grammatical mistakes that the children belonging to the two groups made when answering the questions about the test items were often very similar. This clearly shows that the children in Óbuda have to fight the same struggles as their cousins in Esztergom as far as the development of their language skills is concerned. The active vocabulary of even the older children in the Esztergom group, i.e. of those in the 4th and 5th grades, proved insufficiently small. They quite often lacked the words that they wanted to use to describe the functions of certain objects and their grammatical mistakes made the meaning of their sentences almost impossible to decipher several times. In contrast, the 1st and 2nd grade pupils of the Óbuda school, who had spent as little as two years at school, usually found it easier to express themselves and the grammatical mistakes that their speech contained hampered understanding to a considerably smaller extent.

The answers classified as "absolute errors" mostly reflect notional defects. However, probably, the failure to answer correctly stems from perceptual as opposed to notional difficulties in some of the cases. The type of the answers that the children gave suggests that the child failed to recognize the "meaning" of a picture, in other words, they were unable to identify the drawings consisting of lines, dots and colours with objects or phenomena which they originally met as particular things or objects. For example, one of the 1st grade pupils (Mária S. in the Óbuda group, who had spent one year in school) described the picture of "snow", depicting snow falling and houses with white roofs, in this manner: house, dots. (…) A 1st grade pupil in the Esztergom group described the same picture saying (in the Gypsy language) points.

The children's notional difficulties can be well illustrated by the test data concerning naming different household animals. In almost half of the cases (42.5%) the Esztergom Gypsy children were unable to give the correct names of the rooster, the hen, the duck and the goose that they saw in the pictures in either of the two languages. The answers that the 1st and 2nd grade pupils gave seemed rather shocking in this respect: about one third of the children gave the same one name to three of the four farmyard birds (and sometimes to all the four of them) in Hungarian or in the Gypsy language. E.g. they called the rooster, the hen, the duck and the goose that they saw in the four subsequent pictures khajnyi (hen).

These notional and grammatical difficulties seem to prevail to a large extent in the case of the pupils in the higher grades of the Esztergom school. As compared with the results of the control group's: the results that the 3rd and 4th grade Esztergom children achieved in naming the four birds equalled the results of the 3-4 year-old kindergarten children and the results of the 5th-7th grade Esztergom pupils fell below the level of the results of the 5-6 year old kindergarten children.

Parallel data concerning the Óbuda group of the Gypsy children: in their case only one sixth of the answers concerning the four items demonstrated their ignorance of the names of the farmyard birds. The error rates of naming the four test items in Hungarian in the whole of the Óbuda group was 22.5%, while the error rates of naming them in their mother tongue was just twice as high as that: 45%. (I will come back to this point later.) However, the answers that the children of different grades gave to the same questions demonstrate that within the group of the Óbuda children, it is in fact mainly for the 1st and 2nd grade pupils that these four test items caused considerable notional or (Hungarian) language problems. (As for the absolute error rate, in respect of telling the four test items apart, they achieved about the same results as the 5-6 year old kindergarten children.) It is worth noting that the ten Óbuda working class children in the control group were all able to give the names of the 4 farmyard birds correctly without a single error.

It is remarkable that the rate of errors concerning the Gypsy language answers of the 1st grade pupils belonging to the Óbuda and Esztergom groups (62.5% and 61.1%, respectively) were almost identical. This fact clearly shows that within their own mother tongue environments, the Óbuda and Esztergom children had equally little chance to acquire these notions. (Actually, the adults living in the estate found it natural that the children could not tell the four animals apart in the course of the test. As Bori K. (62), one of the Esztergom grandmothers said: "The little children all say khajnyi (hen), it is a word that children use. Those who know more say different things like "papiny", "ráca", khajnyi" (i.e. goose, duck, hen).

At the same time, it is interesting to note that many of the Óbuda children only new the Hungarian name of the test items that the Esztergom children were unable to give the name of either in the Gypsy language or in Hungarian. The reason why they knew the Hungarian names of these things is that they learnt them at school. In other words, the educational facilities of the Óbuda school - as opposed to those in the Esztergom school - ensured the successful acquisition of the new notions. This is why the Gypsy language rate of errors is twice as high as the Hungarian language rate of errors in the test results of the Óbuda children concerning the test items concerned.

THE CONCLUSIONS OF THE BENDER TEST

The Bender test examines the children's degree of maturity concerning "visual-motor" skills (i.e. the harmony of perception and movement ), using the procedure of reproducing different forms. In the course of the testing, the person tested has to copy 9 complex geometric figures. The figures that they copy are evaluated from these 3 points of view: 1. the rendering of the angles, 2. the directions of the figures and their constituents (orientation) 3. the relative position of the figures and their constituents .

The Bender test (Series B) is suited for testing children between the ages of 6 and 12. However, during the survey, I administered the test among children over 12 as well, since their disadvantages and general achievements made it probable that - at least for some of them - solving the tasks in the test would prove somewhat problematic despite their age.
When administering the test and evaluating the results I used the chart that evaluates the achievements of the children according to their age as well as the methodological manual compiled during the use of the test in Hungary . As for the achievements of the children over 12, I considered the achievements of the children of 12 to be their "appropriate for their age" achievement.

Again, the Bender test showed considerable differences between the results of the two groups. While as few as 8 of the 40 Esztergom children (20%) reached the "appropriate for their age" level, the majority of the Óbuda children, i.e. 12 out of 20 (60%) reached the same level of achievement.

The rate of backwardness of further 8 children belonging to the Esztergom group could be calculated as the equivalent of 1 or 2 years, while the achievements of the majority of these children, i.e. of 24 out of 40 (60%) could be calculated as of 3-4 years or even more below the "appropriate for their age" level. (The children in the Óbuda group who were 1-2 and 3-4 years below the "appropriate for their age" level comprised 20-20% of all the children tested.)

Also, while the average rate of backwardness of the 7-10 year old children in the Esztergom group was only about 2 years, that of the 10-12 year old children was nearly 4 years and that of the children over 12 was nearly 6 years. Considering the fact that this rate of backwardness is typical of the majority of the children belonging to the different age groups (70.5% of the children aged 10-12 and 62.8% of the children over 12 fell in these categories), we can confidently claim that the relative backwardness of the majority of the Esztergom children conspicuously grew over the years. A similar tendency can be noticed in the parallel age groups of the Óbuda pupils, which, however, concerned a much smaller group of the children belonging to the separate age groups, i.e. 40% of the children aged 10-12 and only 22.5% of the children over 12.

Thus we can see that a psychological test consisting of non-verbal tasks used in the survey has the same results as the test of vocabulary: the majority of the children educated in the Gypsy study groups where pupils belonging to different grades studied together were halted in their development and their relative backwardness increased over the years. In contrast, the development of the Óbuda children who studied in separate classes proceeded in a far more favorable manner. The majority of the children were capable of approaching and in certain cases even catching up with the average level appropriate for their age.

THE RESULTS OF THE READING TEST

I used a reading test consisting of 20 tasks (a Hungarian adaptation of a test based on "silent reading", originally devised for testing French children ) to measure the reading skills of the children. The same test had been used for measuring the reading skills of Hungarian 1st grade pupils at several institutions, e.g. at Budapest remedial schools and in the experiment of native language teaching at the Kaposvár Teacher Training College. (The monograph that presents the results of the Kaposvár experiment contains the complete test and the reading skill ratios of the experimental class and the control group. ) We compared the data collected by testing the reading skills of the Gypsy children with the data concerning the average result of these two 1st grade classes.

Each task in the reading test contains a sentence and a picture that belongs to the sentence. The person tested has to read the sentence (silent reading) and modify the picture according to the (hidden) instruction in the sentence - either by adding something or by coloring something. E.g. in Task 10, where the sentence is "I put a flower in the vase", they have to draw a flower in the empty vase in the picture. For each correct solution the pupil is given 1 point.

The results of the reading test in the Esztergom group were shocking. 7 out of the 9 1st grade pupils, 5 out of the 10 2nd grade pupils and 5 out of the 9 3rd year pupils were unable to solve one single task: which means that nearly half (42.5%) of the Esztergom children could not read at all, while in Óbuda, there was only one test that scored 0 point: that of Mária S. (aged 7).

By comparing the results of the Gypsy and non-Gypsy children, evaluated in the same manner, it becomes clear that the average test result of even the 4th grade Esztergom Gypsy pupils (74.2%) is below that of the Kaposvár experiment control group, i.e. the non-Gypsy 1st grade pupils who achieved worse results in the Kaposvár experiment (78.9%). On the other hand, the average test result of the 2nd grade Óbuda pupils (90%) is over that of the pupils of the Kaposvár experimental class, i.e. the "Hungarian" 1st grade pupils who achieved better results in the Kaposvár experiment (85.3%).

THE STANDARDS OF GYPSY CHILDREN'S EDUCATION IN THE TWO SCHOOLS

The significant differences between the test results of the two groups of children of similar socio-cultural backgrounds draws our attention to the standards of education in the two schools.

The unacceptably low achievement of the Esztergom group of Gypsy children suggests that since the beginning of their school career, serious disadvantages had been added to those originating in their circumstances. The factors that contribute to their disadvantages at school are these: pupils of different grades studying in the same classroom, great fluctuation of teachers, extremely badly equipped classroom, total and permanent segregation of children. Also, as a result of all these, as it is reflected in the very high ratio of the 3rd grade pupils still unable to read, another factor is the more or less inevitable abandonment of the imposition of the appropriate minimal requirements concerning each school grade, which are also prescribed by law.

As far as pupils of different grades studying in the same classroom is concerned: children belonging to two/five different grades study together in the two Esztergom study groups while all 700 of the non-Gypsy pupils of the school go to classes of different grades. In the school year 1976/77, one third of the Gypsy children registered in the class logbooks and obliged to attend school according to the regulations did not attend school at all and more than half of the children failed at the end of the school year. Pedagogically speaking, in the year of the survey the educational circumstances secured for the two Gypsy study groups were so adverse that if the children who were regularly absent had attended school, it would have become totally impossible to carry out any schoolwork. The classes and the afternoon study activities of the two groups were held in the same classroom and - according to the information in the logbooks - at the same time on certain days of the week in the school year 1976/77. Had all 64 of the Gypsy children registered in the logbooks turned up on such occasions, there would only had been enough seats for all of them if three or four of the children had been seated at each school-desk for two pupils. (Fortunately, the situation has changed in the present school year. Gypsy children got their classrooms back, and the conditions of the educational process have somewhat improved in other respects, too. Pupils of different grades study in the same classroom and the disadvantages that originate in the permanent segregation of the children, described below, still exist.)

The situation of the pupils in the higher grades of the primary school is extremely controversial, too. In the same educational institution where all non-Gypsy pupils attending the higher grades of the primary school receive education from qualified teachers in each school subject and can study the different subjects in different, specially equipped classrooms, Gypsy children of different grades have a common study group and e.g. during their physics and chemistry classes they only have the chance to imagine the experiments - looking at the illustrations drawn on the blackboard - while their non-Gypsy peers can observe and even administer them in reality in the physics and chemistry classrooms. While their non-Gypsy peers in the higher grades of the primary school are taught Russian by qualified language teachers, Gypsy children do not learn Russian at all. However, I have the impression that the majority of the Gypsy children who get as far as the higher grades of the primary school are rather intelligent and eager to learn, which is not only typical of the schools surveyed. The results of the IQ tests examining 873 Gypsy children by the Department of Hygiene of Children and Young People of the National Public Health Institute clearly show that among Gypsy children, only those with the best abilities get as far as the higher grades of the primary school. Rather typically of the eagerness of the group of the Esztergom pupils of the higher grades, one day they got to surprise their teacher by claiming that they also wanted to learn Russian. (Unluckily, their teacher was unable to take on this task for want of proper command of the Russian language.)

Being a pupil in a Gypsy study group does not only mean disadvantages of an educational nature in the strict sense. It also means that their PE lessons are held in their classroom or in the schoolyard assigned for this purpose instead of the school gym. They have lunch sitting at the uncomfortable classroom desks, the surface of which is slightly slanting, instead of the more civilized school canteen. Basically, as for the relations of the practical daily contacts they are almost totally separated from the other children. At the same time, the circumstances of the Gypsy classes have also failed to provide solutions for the special educational tasks that are very important in the case of Gypsy children (e.g. accustoming children to cleanliness, raising the level of personal hygiene). As for the facilities concerning washing in the school, the 64 children could only avail of one metal washing basin and the cold water in the bathrooms).

In these circumstances, it is not in the least surprising that, as it is reflected in the analysis of the answers of a questionnaire and the description of the situation in a college student's thesis about the problems concerning the education of Gypsy children, the relationship of Gypsy and non-Gypsy children and their comportment towards each other is full of mutual fear and suspicions.

The fact that the isolation of the children is apparently a permanent state of affairs even aggravates the problems: not a single pupil originally placed in a "Gypsy" class has been later directed to attend the "mixed" classes over the years, although the teachers have put forward several recommendations concerning the transfer of certain pupils and some of the parents also asked the school to do so. The only way from the study group of the pupils of the 1st-4th grades leads to another study group for pupils of higher grades, where education is even less directed towards the specific curriculums and requirements of the different grades of primary school.

The results of the Óbuda Gypsy children prove that by implementing certain adequate and purposeful measures, the integration of Gypsy children into the school community can be basically successful. Due to the favorable circumstances the teachers of the school have managed to achieve the following: the majority of the children were able to comply with the compulsory requirements of primary school education, if at a poor level. Only 1 of the 20 children evaluated according to the generally compulsory norms and educated in "mixed" classes (Mária S., aged 7) failed to comply with the minimum requirements at school in the year of the survey.

One of the factors that enhanced their success was surely the school's positive attending to the special educational and pedagogical tasks to be undertaken in the education of Gypsy children. A shower room was installed next to the room where the afternoon study group activity for the Gypsy children was carried out. The Municipality of Budapest offered a position for a social worker assisting Gypsies within the staff of the school. The tasks of this social worker were ensuring the appropriate level of hygiene among the children, assisting their health care, helping their families in all matters that can lead to the enhancement of the circumstances of the children's schooling (arranging help for them in matters of employment and housing, finding places for the younger children in kindergartens, etc.). The children, whose circumstances concerning washing at their homes were not much better than of those living in Esztergom, got washed from top to toe and were made to put on clean clothes at school every day. As a result of the social worker's conscientiousness work, the frequency of the conflicts and problems originating in Gypsy children's poor level of hygiene and the potential of infections has largely diminished.

As for the 1st and 2nd grade pupils' integration, another important aspect was the establishment of the afternoon study group for a small group of pupils, led by a teacher with great experience in coaching and assisting children in afternoon study activities. Meticulous attention was paid to the choice of the right person, whose task was in fact implementing a daily individual coaching programme. The teacher checked the homework of all the 13 pupils studying in 5 different grades on a daily basis, asked them to report orally on what they had learnt and explained to them all that they failed to understand during the lessons. Special emphasis was placed on reading: apart from the texts in the children's readers assigned for homework, they also used the texts in the anthologies of poetry and tales placed on the shelves of the classrooms. Maybe it is mainly due to these efforts that the 2nd grade pupils developed exceptionally good reading skills. (It is also worth noting that no special methods were used - and such methods could not even have been used as special methods for educating Gypsy children are not outlined or developed at all. Simply, this teacher did a good job working within favorable circumstances with children who, during the morning lessons in the mixed classes received education of the same level and intensity as the other pupils attending the same school.)

The free activity sessions, which had a markedly enhancing effect on the children's general aptitude, were successful in the well-equipped, friendly and not at all crowded classrooms. During such sessions, sitting down on the playtime carpet, the children were free to choose what they wanted to play with: sewing clothes for dolls, weaving or reading tales, playing didactic games, using plasticine, drawing, etc. They could use the rooms of the schools used for common activities, i.e. the canteen, the classrooms specially equipped for teaching certain subjects, the pioneers' room, the doors of which used to be closed in front of Gypsy children in the same school in the years before the initiations of the Budapest Pedagogical Institute, when Gypsy pupils had separate study groups of pupils of different grades. The relationship of the Gypsy and non-Gypsy children also took a favorable turn. I witnessed their lighthearted playing together in the break-time between two lessons several times while working on the survey.

Consequently, the results of the survey and the experience gained during the administering of the tests definitely confirm the opinion that questions the rightness of maintaining Gypsy classes as an organizational form. They show that ethnically, linguistically and sociologically, such treatment of the issue that wishes to make the education of Gypsy children in segregated groups permanent is especially unacceptable.

Other people's accounts, which we rely on for want of data, back the impression that the situation of the school in Esztergom is not unique. The person quoted in the introduction - a sociologist participating in carrying out the nation-wide survey of the Gypsy population in Hungary in 1971 - has experienced that the circumstances in Gypsy classes at several places all over the country are very similar to those in Esztergom. I have also witnessed phenomena similar to those described here, i.e. the deteriorating standards of originally well functioning separate study groups in several schools. It is worth noting that the documentary film entitled "Mit csinálnak a cigánygyerekek?" (What are the Gypsy children doing?) presents a similarly controversial picture of the functioning of Gypsy classes.

Even though we have not gathered a sufficient amount of facts to serve for generalization on the basis of the situation in Esztergom, we can claim with all responsibility that the problems described here are typical of the majority of the Gypsy classes and prove the assumptions concerning the dangers and controversy of segregated education and the justify the worries concerning them.

Basically, certain factors of segregated education create a disadvantaged situation at the very beginning of the educational process. It is usually unavoidable to place pupils belonging to different grades within the study group, as there are very few settlements in the country where Gypsy children could fill whole separate classes. In these circumstances, it is not in the least surprising that the policy that directs Gypsy children to attend segregated Gypsy classes often triggers the disapproval of the more demanding Gypsy parents living in orderly circumstances. Although Point 2 of the 18 345/1967 Ministerial Order clearly states that the school needs the parents' approval of placing the child into a separate class or afternoon study group in every case, in practice, it often happens that the school ignores the parents' protests, even in the case of families that have reached quite a high degree of social integration. It has been widely experienced that in schools that run separate Gypsy classes, normal classes do not receive Gypsy children at all or only in the course of a rather cumbersome process.

I assume that it is needless to add anything to prove that this factor - and permanent segregation in general - is and will be an obstacle to Gypsy people's integration into the society.

The success of the children's further school career remains rather uncertain, even if they are transferred to other classes. Pedagogically, the children coming from a segregated study group, where education is less efficient, do not have much chance to achieve among their non-Gypsy classmates if they are not given any assistance. The test results also make it obvious that after a certain period of time it is impossible to transfer the children studying in Gypsy classes similar to the one in Esztergom to normal classes, as the process of their falling behind has become irreversible in the Gypsy classes, originally established to compensate for the backwardness of the pupils.

I am convinced that it would be possible to achieve incomparably better results in the education of Gypsy children if the authorities concerned decided to spend the amounts that the maintenance of permanent Gypsy classes which often bring about discouraging results cost on the organization of intensive pre-school education, health and social care and coaching programmes. The Óbuda experiment definitely proves that this is the practicable way of assisting the integration of Gypsy children.

In whatever way we try to conceive the rise of Gypsy people, we should consider it vital that their disadvantaged circumstances should not be reinforced within the school community. The formation of hundreds of thousands of illiterate and semi-illiterate people within the Gypsy population by the end of the century needs to be prevented. Not only is it the interest of the Gypsy community but also that of the whole Hungarian society.