The Corpus-based Dictionary of Polish Sign Language is based on the Polish Sign Language (PJM) Corpus, which is an extensive collection of recordings of Deaf users of PJM retelling the content of picture-stories and video clips, talking about themselves, and discussing topics of interest to them.
Containing many hours of recorded material elicited from a range of individuals, the Corpus makes it possible to ascertain which PJM signs are used by Deaf signers, and how they are actually used.
As such, the dictionary is based on scientific principles for documenting and describing units of language: it documents and describes real signs spontaneously used by Deaf users of PJM. At the same time, the dictionary may be used by those learning PJM as a foreign language, in other words as a foreign language dictionary.
The definitions which the dictionary offers are written in Polish and are modeled after the definitions in Mirosław Bańko’s monolingual Polish dictionary Inny słownik języka polskiego. They are written in a way that is akin to the kind of definitions to be found in monolingual dictionaries, i.e. they provide semantic information that is more extensive and precise than the sort customarily provided in bilingual dictionaries. The objective of providing such definitions is to enable Polish-speaking learners of PJM to understand the precise meaning of a given sign, and also to enable Deaf users of the dictionary to learn to more effectively and accurately interpret the definitions of meanings given in standard dictionaries of spoken Polish.
The dictionary includes all the PJM signs that were found to appear in the PJM Corpus more than 4 times (as of 2015). However, because not all of the important PJM signs happened to occur in the recordings (not having been required by the story-telling or other tasks, and also not having been used by signers in free discussion), we decided to fill in the obvious gaps in the list so compiled.
To this end, we classified the signs into semantic categories and then added the basic signs that were missing from each category. Note that this categorization did not encompass all signs, as many are of such universal meaning that they cannot be assigned to any of the categories. Similarly, even when a sign occurred many times in the corpus, not all of its senses were necessarily represented there. Our Deaf and CODA editors, in consultation with other Deaf individuals, sometimes decided to include additional senses of individual signs.
At first glance it may seem that the dictionary covers only a small number of signs. This is a result of the decision that every sign will be represented in the dictionary only once, even if it has a large number of unrelated senses. Moreover, it is usually the case that one PJM sign corresponds to a whole series of interrelated words in spoken Polish. For example, a single PJM sign corresponds to the Polish words śmieszny ‘funny’, zabawny ‘fun’, (dobry) humor ‘(good) humor’, kawał ‘joke’, dowcip ‘jest’, żart ‘jape’. Nevertheless, when giving Polish equivalents we did not list a whole series of such closely related expressions, but rather tried to find one expression that most accurately corresponded with the meaning described.
Lexicographic work on spoken languages, particularly Polish lexicography, traditionally draws a clear distinction between cases of homonymy and polysemy. Homonyms are words that are coincidentally written and pronounced in the same way: often these are independently borrowed expressions, frequently originating from different languages. Spoken Polish homonyms, for instance, include the preposition bez ‘without’ and the noun bez ‘lilac’. Such items are traditionally given separate dictionary entries, often under successive numbers, e.g. bez1 and bez2.
In cases of polysemy, in turn, a word has various meanings that bear some relation to one another. For instance the Polish word krem ‘cream’ may refer to a sweet desert (ciastko z kremem ‘cream-filled cake’) or to a thick soup of uniform consistency (krem ze szparagów ‘cream of asparagus’) or to a cosmetic product rubbed into the skin (krem przeciwzmarszczkowy ‘anti-wrinkle cream’). What all these senses share is the notion of a uniform substance that can be easily spread. On the other hand, these are all distinctive senses of the word krem, not examples of the word’s use in a single sense with respect to different things. It is not the case that krem describes any uniform, easily spreadable substance: for instance, such a substance used to protect machine parts will be called smar ‘grease’ instead. In dictionaries of spoken Polish, a number of criteria are usually applied simultaneously to distinguish between cases of homonymy and polysemy; these include independent origins, no shared meaning, belonging to different parts of speech, etc.
In this dictionary of PJM, we decided not to introduce such a distinction for both technical and substantive reasons.
Firstly, there is no analogous way to ‘add’ a distinguishing numeral to a sign-language sign; this means that the different meanings of a sign would still have to be listed together in the same entry, under the same headsign, irrespective of whether the sign is considered homonymous or polysemous – the distinction could only be drawn lower, within the entry itself. Secondly, because of the iconic origins of many sign-language signs, it is very difficult to decide whether we are faced with two signs that are performed in exactly the same way, but have meanings that are completely unrelated, or a single sign with multiple meanings. The iconic nature of sign-language signs also makes it difficult to speak of coincidental correspondence, even when unrelated meanings are represented by a single sign.
For these reasons, the Corpus-based Dictionary of Polish Sign Language does not distinguish between homonymous vs. polysemous expressions. In all cases when the meanings of a given sign fall into groups of senses, some of which are closely related, the entry is divided into what are called ‘macrosenses’. Macrosenses may be completely independent of one another – consider for instance the PJM sign that can mean both ‘profile’ (macro-sense I) and ‘Portugal, Portuguese person, Portuguese’ (macrosense II) [http://www.slownikpjm.uw.edu.pl/gloss/view/1253].
They may also be related in some way, as in the case of the PJM sign that can mean ‘a lot’ (macrosense I) and ‘sufficient, enough’ (macrosense II) [http://www.slownikpjm.uw.edu.pl/gloss/view/925]. The division of meanings into macrosenses was performed based on the linguistic intuitions of the lexicographers.
Macrosenses are distinguished by successive Roman numerals: I, II, III… If an entry is not divided into macrosenses, that means all of its meanings are interrelated within a single macrosense.
When describing PJM, we essentially cannot speak about the same ‘parts of speech’ that are familiar from spoken Polish, where nouns, verbs, and adjectives very distinctly differ from one another in morphological form, even if they may be closely related in meaning as a result of derivation. In PJM, the same sign may be used in a way that is analogous to a Polish verb (as the predicate of a sentence), or in a way analogous to a Polish noun (as a subject or object). Another sign might be used one time as a noun (for instance in the meaning ‘Portugal’), but another time adjectivally, for instance when talking about ‘Portuguese seafood’. The situation is a bit similar to languages like English, where parts of speech are not as heavily morphologically marked as in Polish: for instance, the bare infinitives of a large number of English verbs, such as sledge, are identical in form to the related nouns.
The meaning of a sign-language sign very often depends on whether it is used as an adjective, or as a noun or verb. For instance, the PJM sign corresponding to the Polish words bać się ‘be afraid’ or tchórzliwy ‘cowardly’, when used as a verb means that someone is scared of something, whereas as an adjective it means that someone has a certain proclivity for being scared. The difference here is clear-cut: someone who is generally brave may feel afraid in a concrete situation, whereas someone who is generally cowardly may show exceptional bravery in some specific situation.
To capture these differences, the portion of an entry dealing with a specific macrosense may be further divided into types of use, denoted with capital letters: A, B, C…
Three large classes of types of use have been distinguished: the first consists of signs used as fully-denoting expressions, in other words as referring to something in the world. In their meanings, these uses are akin to fully-denoting expressions in many spoken languages: i.e. nouns, verbs, adjectives, adverbs. These expressions and signs may be thought of as the bricks used to build a sentence. The dictionary therefore marks these uses with annotations such as ‘in nominal use’, ‘in verbal use’, ‘in adjectival use’, ‘in adverbial use’, ‘in numerical use’, and the corresponding definitions contain the phrase ‘refers to…’ (odnosi się do…).
However, there are signs that do not refer to anything in the world, but rather link fully-denoting expressions together into an sentence. Such lexemes are known as function words. To continue to above metaphor, they are something like the mortar that holds together the bricks, making the sentence a coherent whole. In spoken languages like Polish and English, these are conjunctions, prepositions, pronouns, without which sentences would be incomprehensible. In sign languages, in turn, there is usually less such ‘mortar’ than in spoken languages, because the same role is partly played by the signing space, mimicry, body position, head position, gaze direction, and changes in body position, head position, and gaze direction. Nevertheless, there are signs in PJM that do act as such mortar. The dictionary therefore marks uses of this second kind with annotations such as ‘in pronominal use’, ‘in conjunctional use’, ‘in prepositional use,’ ‘in particle use’, and the corresponding definitions contain the phrase ‘used to…’ (służy do…). Fully-denoting signs (in uses corresponding to the traditional fully-denoting parts of speech) and functional signs combine together to form sentences.
There is also a third kind of signs in PJM, ones that can only occur independently. In this regard they are similar to spoken Polish interjections, such as Aha! or Brawo! Like spoken interjections, such sign-language signs do not combine together with other lexemes to form sentences. These signs are described as being used autonomously (i.e. outside of a sentence). Sometimes their equivalents in spoken Polish are not single words, but whole phraseological expressions.
Some signs used autonomously have a very particular function. Some of them are clearly used to express one’s attitude towards one’s own statement, or another statement made by the interlocutor. Such cases are described as meta-textual use. For instance, the sign that means ‘one’ when used as a numeral, when used meta-textually ‘expresses a positive response or acceptance’ (with respect to a question or proposal from the interlocutor) [http://www.slownikpjm.uw.edu.pl/gloss/view/141].
And so, the annotations concerning types of use should be interpreted as follows:
Because the dictionary introduces a very detailed division into types of uses, when separating the different senses of signs we avoided the kind of very detailed divisions that are characteristic of dictionaries of spoken Polish. This difference is in large part due to the nature of sign-language signs themselves. On the one hand, their meanings are very frequently more general than the meanings of Polish words. For example, corresponding to the Polish words wiadukt ‘viaduct’, most ‘bridge’, and kładka ‘footbridge’ is just a single PJM sign, referring to a structure making it possible to walk or drive across a river, roadway, railway line, or other obstacle. On the other hand, because of their iconic nature, sign-language signs may also have a significantly narrower range of meaning than Polish expressions.
This means that even if a sign-language sign corresponds to a whole series of Polish expressions with similar, though not identical meanings, that does not mean that the sign itself is polysemous: on the contrary, its meaning often corresponds to that portion of the meaning that is shared by the near-synonymous expressions in Polish, with the differences between the Polish expressions being irrelevant to the meaning of the sign-language sign.
Typical examples of this phenomenon can be found in the names of professions, functions, and certain kinship relations. For example, Polish has the expressions brat ‘brother’ and siostra ‘sister’, carrying information about gender, whereas PJM has a single sign meaning ‘child of the same parents’. The same holds for other gender-marked pairs of words, such as uczeń/uczennica ‘(male/female) pupil’, nauczyciel/nauczycielka ‘(male/female) teacher’, lekarz/lekarka ‘(male/female) doctor’, pielęgniarz/pielęgniarka ‘(male/female) nurse’ etc.
The field of lexicography and lexical semantics (the part of linguistics that studies the meaning of expressions) has developed special tests by which one can determine whether we are dealing with a single, broader meaning, or with several close but different meanings. We applied these tests with the assistance of our Deaf consultants.
In the Corpus-based Dictionary of Polish Sign Language we decided to provide so-called descriptive definitions, which are characteristic of monolingual dictionaries. What makes our dictionary different from such monolingual dictionaries is the fact that although the headwords of the entries are PJM signs, the definitions have been formulated in Polish. Otherwise, the dictionary would not be usable by individuals who are learning PJM.
There are many possible ways of defining expressions (or signs). Traditional dictionaries of spoken languages used so-called ‘encyclopedic definitions’ – invoking scientific knowledge, as represented for instance in encyclopedias. Modern lexicographical practice reserves such definitions for true scientific terms, whereas general-use expressions are defined in a way that relates to everyday knowledge about the world. In the Corpus-based Dictionary of Polish Sign Language, we have used such non-technical definitions for all meanings relating to generally-known phenomena (the PJM signs covered are predominantly of this sort).
The structure of the definitions and the information they contain depend on the type of use. For fully-denoting signs, we have tried to provide what are called referential definitions. They describe the thing, phenomenon, property, or situation that the given sign refers to.
The definitions of signs used nominally provide a description of what is referred to. This description contains certain characteristic properties of the object: its appearance, size, purpose, sometimes taste or smell – these are properties closely associated with the given object. For example, the meaning of a sign that may refer to two vegetables, the carrot (1) and the parsley root (2), was defined as follows (here cited in English translation):
1. <<refers to the edible root of a plant, taking the form of an elongated cone, usually with a distinctive orange color, added to soups, eaten boiled or shredded raw>>
2. <<refers to a plant with a white elongated root, but also its green leaves, used as an addition to soups and dishes>>
For signs used adjectivally, the definitions provide not only a description of the property specified by the given sign, but also a description of the person or thing which may be described as having that property. For instance, the sign corresponding to the Polish adjective śmieszny ‘funny’ is defined as follows (again, in English translation): <<refers to an individual or their actions which provoke laughter>>.
In the case of signs used adverbially, the definition usually describes properties of the situation, activity, or action characterized by the given sign. This provides information about which kinds of signs used verbally the sign being defined can combine with: those which describe activities, actions (taken by people), or more broadly any situations in which something happens (for instance, the definition for the sign corresponding to the Polish word długo ‘for a long time’ states <<characterizes a situation in which someone does something, such as talking, for a large extent of time>>).
Understandably, the most extensive definitions are for meanings related to verbal uses. Such definitions specify whether the sign refers to an action (usually physical), to a more general activity (more complex, conscious, not necessarily exclusively physical), or to an even more general situation. Actions and activities are defined in terms of what is done, sometimes by describing the objective. Moreover, definitions mention the participants of a situation: a person (described as ktoś ‘someone’ or jakaś osoba ‘a certain person’) or thing (coś ‘something’). The person or device so described indicates what can be the grammatical subject of a sentence using the sign being defined, in the meaning being described. For example, the sign which (in a specific use and meaning), corresponds to the Polish verb jeść is defined in the Corpus-based Dictionary of Polish Sign Language as (in English translation) <<refers to the action in which someone consumes a meal by chewing and swallowing food>>, and the sign corresponding to Polish obudzić się ‘wake up’ is defined as <<refers to a situation in which someone comes out of a phase of sleeping and opens the eyes>>.
Definitions also indicate whether another individual, animal, or thing is subject to the given action or activity – the potential grammatical objects in a sentence containing the sign in the meaning being described. This can be seen in the definitions of the signs corresponding to the Polish verbs gotować (<<refers to the action in which someone prepares a warm meal or brings it to a boil>>) and pomagać (<<refers to a situation in which someone supports someone in various activities>>).
The same principle applies to the definition of meanings referring to more abstract situations in which something happens, but we cannot speak of an action or activity.
As we have already mentioned, this way of stating definitions is modeled after the definitions given in the monolingual dictionary Inny słownik języka polskiego edited by Mirosław Bańko.
For signs of a functional nature, definitions provide information about the role they play in forming an utterance. This information is provided using the formula ‘used to…’ (służy do…), as in the following definition of a sign:
A. in pronominal use:
1. <<used to ask questions about a reason or consequence>>
czemu ‘how come’, dlaczego ‘why’
2. <<used to ask questions about purpose >>
po co ‘what for’
B. In conjunctional use:
1. <<used to combine clauses to express a cause and effect relation>>
dlatego ‘that is why, for that reason’
Similarly, in the case of autonomous (meta-textual and pragmatic) uses, definitions provide information about the communicative function of the given sign, as for instance in these definitions (here in English translation):
in autonomous use: <<used to express that the undesirable state of affairs under discussion cannot be changed and must be reconciled with>>
in pragmatic use: <<used to urge the stopping of some activity>>
dosyć, dość ‘(that’s) enough’
in meta-textual use: <<used to signal that what is being said is intended as an illustration of the issue discussed>>
The examples in the Corpus-based Dictionary of Polish Sign Language are drawn from authentic signed utterances found in the PJM Corpus. In order to standardize their appearance the original utterances were re-recorded by Deaf members of the dictionary team.
Examples are not provided for all types of use. Given the properties of sign language, not all authentic utterances would be understandable without a broader context. For instance, many of the recorded utterances are fragments of stories, signed using a signing space set up at an early stage of narration. Also, utterances containing pointing signs referring to previously introduced elements, or classifier predicates would be similarly incomprehensible without a broader context.
In terms of their structure, PJM utterances are more similar to colloquial verbal utterances in spoken Polish than to written Polish: there is a prevalence of elliptical utterances, referring back to earlier utterances.
In the Corpus-based Dictionary of Polish Sign Language, examples are provided at the macrosense level, not at the type-of-use level: not for all types of use could examples be found in the corpus, and in specific utterances from the corpus it was not always possible to clearly determine which of the possible types of use was exhibited.
The entries for certain signs may list links to other entries, with two possible motivations. Firstly, linked entries may draw attention to signs that appear similar, often differing in terms of just a single parameter, and may therefore be potentially misleading for learners of PJM. Secondly, the links may point out synonymous signs, or signs partly similar in meaning.
Below any linked entries, there appears a formal description of the sign. This includes:
For each parameter, there is an attached link ‘see all’ (zobacz wszystkie), enabling the user to find all the dictionary entries with the same parameter value.
A selected sign can be flagged as ‘to learn’ (do nauki). A list of those so flagged can then be referred to.
There is also a function of printing out an individual entry (in black and white).
The dictionary interface offers the ability to search for a sign according to its form and by certain semantic properties (advanced search).
The form of a sign is defined by three parameters:
Basic handshapes are defined in terms of PJM fingerspelling signs (A, B, C etc.). Numbers added to a letter refer to a modification of the fingerspelling sign. Numbers 1-3 refer to thumb position, 4-7 to the position of the remaining fingers, and 8-9 to other modifications:
1 – thumb perpendicular to other fingers,
2 – thumb bent into the palm/hand,
3 – thumb extended in front of the palm/hand,
4 – fingers bent at acute angle with respect to the palm/hand,
5 – fingers bent at right angle with respect to the palm/hand,
6 – fingers bent at obtuse angle with respect to the palm/hand,
7 – fingers arched.
The handshape parameter is selected by clicking on the respective picture.
The localization parameter shows the location of a sign on the signer’s body. Some signs are positioned on the other hand, and the description is with respect to that hand. The respective options are shown in the pictures with captions; the localization parameter is selected by clicking on the respective picture.
Additional parameters include whether a sign is two-handed, performed close to the body, with touch, or a change in hand position as the sign is performed.
Advanced searching is based on the semantic properties of signs. These include:
Certain signs, at least in some of their meanings and types of use, clearly belong to a certain thematic field, such as education, health, body parts, etc. However, not all fully-denoting signs can be assigned to such a semantic category.
The dictionary makes it possible to search for signs related to the individual categories. Of course, not all the senses of a given sign have to fall within the given category. A category is selected from a list.
The user can also search for signs based on types of use. This is useful in finding function signs, or in finding fully-denoting signs within a given thematic category. For example, we can search for a sign referring to some kind of evaluative attitude that can be used as a verb. Of course, in such a case both verbal use and referring to an evaluation do not have to be properties of the same sense of the sign. For instance, for the following sign http://www.slownikpjm.uw.edu.pl/gloss/view/217 only the meaning ‘to stare’ involves such an evaluation.
Searches can also be carried out based on Polish words given as equivalents. For a sign to be found in this way, the Polish word must actually be listed as one of the sign’s Polish equivalent. As a result, a search on marchew ‘carrot’ will yield no results, because the corresponding sign lists the alternate (diminutive) form marchewka.
The two search modes (simple and advanced) work together. This should be borne in mind, and if necessary the previous query should be cleared.
The dictionary interface also offers the ability to flag signs which the user wishes to come back to, for instance when learning PJM.
Clicking the field “to learn” (do nauki) will put the selected sign on the list of signs to be learned. From this list, one can move directly to the entries for the signs listed.
Every entry of the Corpus-based Dictionary of Polish Sign Language is accompanied by a transcription in the Hamburg Sign Language Notation System (HamNoSys). HamNoSys may be viewed as a sign language equivalent of the International Phonetic Alphabet (IPA) used for transcribing spoken languages. HamNoSys is primarily a research tool, and not a common writing system (as opposed to the SignWriting system, for instance).
HamNoSys was created in 1985 by a team of researchers from the University of Hamburg (led by Thomas Hanke) and has since been systematically developed. The current version is 4.0.
The HamNoSys transcription is designed to be universal, i.e. it can be used to transcribe signs of any sign language. It takes into account all possible handshapes, localizations and types of movement. Different articulation parameters are represented as symbols and numbers.
In the Corpus-based Dictionary of Polish Sign Language, HamNoSys is used to transcribe the basic form of each lexeme, i.e. the citation form. An example of this kind of transcription is shown in the figure below.
The above symbols correspond to the following parameters: handshape, hand orientation (composed of finger orientation and palm orientation), sign localization and movement involved.
Fragments of this introduction are based on the following paper: Jadwiga Linde-Usiekniewicz, Paweł Rutkowski, The division into parts of speech in the Corpus-based Dictionary of Polish Sign Language, In: Proceedings of the XVII EURALEX International Congress, Tbilisi, 2016.
Joanna Łacheta, Małgorzata Czajkowska-Kisil, Jadwiga Linde-Usiekniewicz, Paweł Rutkowski (eds.), 2016, Korpusowy słownik polskiego języka migowego/Corpus-based Dictionary of Polish Sign Language, Warsaw: Faculty of Polish Studies, University of Warsaw, ISBN: 978-83-64111-49-5 (online publication).