Surces of information on functioning of phonological system


Bronisław Rocławski



            The controversiality of the phonological research results is caused by the fact that the methodology of this kind of research is quite arbitrary and there is no exact reference between phonetics and phonology. Moreover, the linguistic awareness among users of a particular language has not been appreciated (or even disregarded) while evaluating the linguistic units and structures.

            The duty of the researcher of phonological phenomena (phonologist) is to determine and use all of the sources that bring information on functioning of the phonological system. Other important tasks include establishing the priorities for certain information sources and the algorithm of methodological procedure ensuring that the source data is verified in a relatively objective manner. These tasks are very complex. We can present here only some of the problems in more detail because of space limitations.

            We consider that for the phonological research results, it is an extremely important question to deliberate the aim of particular methodological operations itself. It should be also noted that the methodological procedure itself will also depend on the information sources and the attitude of the researcher towards the source material. Discussing a specific methodological procedure, we assume that the phonologist is a user of the given language. Thus, the language under consideration is a living and modern one. In its written form, this language has an alphabetic writing system.

            The methodological procedure proposed in this work is aimed at establishing such a phonological system that will serve the didactic purposes in the best possible way. We believe that such a phonological system is most similar (or identical) to the one actually functioning in the given language. This goal is different from the one set by the researchers who focus on determining a stock of phonemes used to automatically recognize spoken texts and to automatically create them.[1] Our aim coincides with the one set by the grammarians since many generations. Linguistics is an applied science. The results of linguistic research are used, above all, to teach languages. In the process of language acquisition, there are living language users, thinking members of a particular language community.

            The shortage of linguistic research is due to, among others, a false belief that linguistic facts can be explored without language users taking part in it, without referring to their linguistic awareness. Such attitude was typical of authors of phonology. In his Grounds of phonology (Podstawy fonologii), N. S. Trubiecki writes:


            While defining the phoneme, any reference to the “linguistic awareness” should be excluded. It is for the fact that “linguistic awareness” is either a metaphorical expression of a language system (langue) or an utterly fuzzy concept which needs a definition itself, or, possibly, cannot be defined at all.


            In his Course in modern linguistics, Ch. F. Hockett expresses a similar opinion. A. Martinet also rejects the language intuition in the phonological analysis. It is proposed by some few researchers only that the language intuition should be assumed as a criterion used for establishing the stock of phonemes.[2]

            Rejecting the linguistic awareness while determining the stock of phonemes, the phonologists have to look for formal rules of determining phonemes. These differentiate depending on various phonological schools. W. Mańczak strongly criticizes the rules elaborated by N. S. Trubiecki:


            ... these rules are not useful at all, and it allows us to understand, why many other phonologists after Trubiecki tried to solve this problem. However, their efforts fell through as well, and it was for the simple fact that the concept of establishing the rules, which would allow to decide if there is one or two phonic sounds in a given case, is as impossible to solve as the problem of perpetuum mobile or squaring the circle. In this case the phonologists forget that all linguistic phenomena are also social phenomena, and social phenomena differ from natural phenomena in their conventional nature. (...) The idea that an objective method of isolation of language units, which are called phonological system by the phonologists, can be elaborated, without making inquiries among people speaking the given language as native speakers, is a daydream that will never come true.[3]


            Despite great efforts of many researchers, such rules have yet not been established that, used by different researchers, would give the same result. Looking at the results of phonological research conducted even in Polish language, a pessimistic conclusion can be made that the number of phonological systems within a given language grows along with the number of researchers working on determining the stock of phonemes.[4] The degree to which a given system is generalized depends mostly on the scientific authority of its author. The phonologists do not have an arbiter to decide a dispute.

            In the methodology of the phonological research that we propose, the linguistic awareness of language users, who are not burdened with reading and writing skills and whose language is developed on a proper stage and level, becomes an arbiter deciding phonological disputes. Such arbiters can be 5-7 year old children. This judgment has been formulated on the basis of tests that have been conducted in the field of phonological awareness in infants.[5]

            If an infant is to act as an arbiter deciding phonological disputes or his/her own phonological doubts, it should be determined whether this infant is a “linguistically adult” human. In Course in modern linguistics, Ch. F. Hockett states:


                A normal 4-6 year old child is a linguistically adult human being. Apart from some unimportant exceptions, he or she uses the phonological system of his/her language, deals with the grammatical substance without efforts, knows and uses basic vocabulary.


            This opinion is not shared by all linguists. In Linguistic grounds of methods of foreign languages teaching, L. Zabrocki writes that a child possesses an almost ready phonological system only at the age of fourteen. Our pedagogic-linguistic research carried out through several years in different environments confirm the above mentioned opinion of Ch. F. Hockett. Raised in normal linguistic conditions, a 5-6 year old child is well prepared to be tested with regard to the functioning of the phonological system. The source of basic information on phonological units is the phonological segmentation of texts. Children are taught the phonic sound (phoneme) segmentation in phonic sound analysis and synthesis classes in infant schools. These kind of classes are aimed at improving the process of writing and reading learning.[6] Both educators and psychologists believe that these tasks are not beyond intellectual abilities of children.

            The process of segmentation is disturbed in case of orthophonic disorders or occurrence in words of phonic sound groups which are difficult to segment. It can be observed that infants often hesitate to pronounce nasal vowels. Thus, children with a good pronunciation should be chosen for the tests. The exercises related to the phonic sound analysis and synthesis should be conducted in the form of didactic games. The best results are obtained when the analysis and synthesis are started with syllables. Next, phonic sound division is practiced using short words with a clear structure containing good contrasting phonic sounds (phonemes). Children become familiar with the phonic sound division of words containing extendable phonic sounds quite quickly. Words containing phonemes which will be the subject of our observation should be avoided during exercises of phonic sound analysis and synthesis techniques. We are not allowed to suggest our own solutions.

            The whole stock of the phonemes of a given language can be established in the process of the phoneme segmentation. Though, we know that it would be an unnecessary action since many phonemes in well-known languages do not raise any controversy. Thus, in our research we focus on controversial phonemes (i  y  ę  Ǫ  į  ų  ń  ŋ  m’  f’  v’  k’  g’  p’  b’ belong to such phonemes in Polish).

            The main objection, which can be faced by our methodological procedure, will concern the assumption adopted here that an infant possesses a fully shaped phonological system. We should look for adult informers in order to avoid this objection. The number of informers who are not burdened with the knowledge of orthographic form of texts is limited to a minimum because of the fact that writing and techniques of reading are so widespread. However, we know that presently the degree to which the Polish orthography has been mastered evokes many critical comments.[7] Even Polish philology graduates make orthographic mistakes. We believe that orthographic mistakes are the source of information on functioning of the phonological system and on the phonological awareness of a given language user. We can limit our observation to the mistakes caused by discrepancies between the phonological form of a text and its orthographic notation only. Interesting material can be obtained e.g. from the observation of incorrect notation of words containing the following letters: ę  i  ą. The material obtained in this way is used above all to determine the place of occurrence of certain phonemes. For Polish, the occurrence of nasal vowel phonemes /ę/ and /Ǫ/ in particular positions is a matter of argument. Frequent mistakes in the notation of ę and ą in the position before the labio-plosive sounds p and b have been confirmed during observation from the research of phonological segmentation which showed that in the modern Polish language there is a two-phoneme group in this place. Orthographic mistakes can be a source of information on distribution changes of certain phonemes that happen in the language. The fact that presently the phoneme /Ǫ/ is frequently noted as a two-letter group oł or ął at the end of words proves that there is a tendency to replace the above phoneme by a two-phoneme group.

            Another source of information on functioning of the phonological system will be the data obtained from the observation of foreigners learning a given language as their second (next one) language. However, a thorough knowledge from the field of contrastive phonology is required to use this source. Intensive contrastive tests conducted nowadays will allow to use this source in a better way in the future phonological research.

            Three sources of information on functioning of the phonological system mentioned above rely on the linguistic awareness of language users. Drawing information from these sources, we established a phonological system of the modern Polish language which contains 38 phonemes (excluding the zero phoneme): i  y  e  a  o  u  ę  Ǫ  į  ų  r  l  m  n  ń  f  v  s  z  ś  ź  š  ž  x  c  з  ć  з’  č  зˇ  k’  g’  p  b  t  d  k  g.[8] Two affricates phonemes /k/ and /g/ have the weakest position in the language. Nasal vowel phonemes /ę/ and /Ǫ/ inhere in the system in a relatively strong way, however, there have been some tendencies concentrated them around for a long time that threaten their phonological independence.

            The phonology should use the data from linguistic research. It is a traditional source of information. The observation of pedagogic-phonetic phenomena is especially important in our research. A comprehensive phonological research should precede the phonological segmentation tests in children. It is necessary in order to evaluate the level of pronunciation skills. The material from two sources must be available, otherwise it is not possible to compare the pronunciation. We must have the results of orthophonic research conducted in the environment of adult informers at our disposal, since we assess the orthophonic “maturity” of infant children. While determining the phonetic structure, we use the data of auditory, articulatory, acoustic and visual phonetics.

            The phonological research requires information on the degree of phonological and phonetic hearing skills. For phonological hearing research, we use tests based on certain preliminary phonological and semantic observation. We confront the words of different meaning and mutually contrasting in a possibly smallest number of phonological elements (one phoneme). At the age of 5, normally developed children posses a well-formed phonetic hearing which can be connected with an ear for music. The state of phonetic hearing conditions the quality of orthophonic occurrences and, to a low degree, affects the phonological segmentation of texts. Informers with a normal physiological hearing should participate in phonetic-phonological tests conducted for the purposes of phonology.

            Results of phonotactic research (here also theory/informative) can be also used in phonological research which are aimed at determining the stock of phonemes. The results of research from these fields are especially valuable when the phonetic units (segments, phonic sounds) are taken into consideration during the tests. However, not much research of this kind has been conducted so far. Above all, the frequency and connectability of phonemes have been examined. Though, this type of research requires the assumption of a certain phonological system. The results obtained can, to a certain degree, serve the future research of functioning of the phonological system. They indicate which places in the phonological system require a special observation. They indicate which places in the system are subject to different tendencies that change their phonological status. In most cases, controversial phonemes have a disturbed connectability and a relatively low frequency. Let us have a closer look at the data on connectability and frequency of Polish phonemes. The data in the table presents the back and front connectability field and the importance of phonemes.[9] Many phonemes that presently raise controversy are at the end of the ranking (Ǫ – 32, k – 34, ę – 35, g’ – 38). They also have a clearly disturbed connectability. The back connectability field is at least two times bigger than the front field (k – 25 : з, g’ – 15 : 3).

            The above deliberations show that in our phonological research we use various sources which, to a certain degree, deepen our knowledge on functioning of the phonological system.


Frequency and connectability of Polish phonemes







Częstość %

Frequency %

Pole łączliwości

Connectability field






[1] This subject has been approached by P. Łobacz, in his article Entropy and acoustic parameters as criteria of phonetic interpretation (Entropia oraz parametry akustyczne jako kryteria interpretacji fonetycznej), Bulletin of PTJ (Polish Lingustic Society), 29.

[2] B. Malmberg writes about it in his New trends in linguistics (p. 148 of the Polish translation): “Some linguists (e.g. Hjalmar Lindroth from Sweden and Amado Alonso from Spain) have proposed that the language intuition of the speaker should be a criterion on basis of which the stock of phonemes of a given language can be determined. Such conclusion can be practical and it often leads to achievement of proper results. Nevertheless, it results in some difficulties of theoretical nature. The language intuition is different in different persons and, moreover, it is influenced by various external factors. Thus, it cannot constitute the basis for determining the phonemes. (...) Thus, later structuralists generally preferred to choose more objective criteria to determine phonemes of a language as well as more formal definitions of the phoneme notion, based on e.g. the origin and distribution of the phoneme, i.e. its possibility to be connected with other elements. Thanks to that, they managed to avoid assuming a psychological point of view (language intuition etc.) and a physical substance (articulation, sound waves, thanks to which the phonemes manifest in speech), which was not the case by the Prague phonologists.” N. van Wijk defined the phoneme as the smallest unit which is considered indivisible by the linguistic awareness. See also remarks made by G. Lindner in his Introduction to experimental phonetics, p. 35 of the Polish translation.

[3] W. Mańczak, Some problems of general linguistics (Z zagadnień językoznawstwa ogólnego), Wrocław 1970, p. 192.

[4] In Course in modern linguistics, Ch. F. Hockett writes: “We should not wonder that different researchers working on the same languages sometimes have different opinions; it wonders and encourages at the same time that, in general, these differences refer to detail.”

[5] The research was started in 1974. I shared the initial results with the participants of the Polish Linguistic Society Convention in Łódź in 1976, in the paper titled On methodology of phonological research.

[6] I present these questions in detail in Phonetic handbook for teachers of infant schools and lower classes of primary school (Poradnik fonetyczny dla nauczycieli przedszkoli i niższych klas szkoły podstawowej) which will be published by Wydawnictwa Szkolne i Pedagogiczne publishing house in 1980.

[7] B. Rocławski On the need to reform the Polish orthography, Scientific Journal of Faculty of Humanities of University of Gdańsk. University of Gdańsk, Linguistics Works 4, Gdańsk 1976, p. 133-139.

[8] This system is presented in my work titled Outline of phonology, phonetics, phononotactics and phonostatics of modern Polish, Gdańsk 1976.

[9] The data comes from the most recent phonostatic tests which I have conducted on the material consisting of 69169 words of spoken and written Polish language.