UNIWERSYTET IM. ADAMA MICKIEWICZA WYDZIAŁ NEOFILOLOGII INSTYTUT JĘZYKOZNAWSTWA TOWARDS A LINGUISTIC WORLDVIEW FOR ARTIFICIAL LANGUAGES IDA STRIA Rozprawa doktorska napisana pod kierunkiem prof. UAM dr hab. Ilony Koutny Poznań 2015 Acknowledgements First and foremost, I would like to express my deepest gratitude to my both supervisors: the initial one, Professor Jerzy Pogonowski, who guided my interest towards artificial languages, for his patience and leniency, and the final one, Professor Ilona Koutny, to whom I am truly indebted for her relentless support and belief in me. No short note can express how great her help was. I would also like to thank my colleagues from the Institute of Linguistics, particularly from the Department of Applied Logic, for their invaluable tips and encouragement (special thanks to Sławek and Agnieszka). My thanks also go to Michael Farris for putting up with my English. Finally, I wish to thank my husband for his love. Q: How many Lojbanists does it take to change a broken light bulb? A: Two: one to decide what to change it into and one to figure out what kind of bulb emits broken light. All errors are my own. Table of Contents Acknowledgements ............................................................................................... 2 List of figures ........................................................................................................... 4 List of tables............................................................................................................. 5 Introduction ............................................................................................................ 6 Research objectives................................................................................................................... 6 Pre-existing research................................................................................................................ 7 Overview ....................................................................................................................................... 8 1 Linguistic worldview ................................................................................. 9 1.1 Short history ................................................................................................................... 9 1.2 Definitions and crucial concepts .......................................................................... 12 1.3 The relation language – culture – thought ........................................................ 17 1.4 Methodology ................................................................................................................ 20 2 Artificial languages ................................................................................. 29 2.1 Definitions ..................................................................................................................... 29 2.2 History and motives of creation ........................................................................... 34 2.2.1 Universal languages .............................................................................................. 38 2.2.2 Formal languages ................................................................................................... 48 2.2.3 International auxiliary languages ................................................................... 50 2.2.4 Artistic languages .................................................................................................. 59 2.2.5 Modern glossopoeia .............................................................................................. 66 2.3 Typology and classifications .................................................................................. 69 2.3.1 Traditional classifications: structure and source of material.............. 70 2.3.2 Traditional classifications: purpose ............................................................... 72 2.3.3 Blanke’s functional classification .................................................................... 78 3 Natural vs. artificial ................................................................................ 81 3.1 Properties of human language .............................................................................. 81 3.2 The scale of naturalness .......................................................................................... 89 3.2.1 The continuum of deliberate influence ......................................................... 89 3.2.2 Lyons’s classification ............................................................................................ 94 3.3 Borderline cases ......................................................................................................... 99 3.3.1 Pidgins and creoles ............................................................................................... 99 3.3.2 Language revitalisation and revival ............................................................ 103 3.3.3 Other cases ............................................................................................................ 106 4 Esperanto as a transitional case ...................................................... 112 4.1 Naturalness of Esperanto ..................................................................................... 112 4.2 Sociolinguistic situation of Esperanto speakers ......................................... 114 5 Linguistic worldview for artificial languages ............................. 125 5.1 The subject: author or community? ................................................................. 125 5.2 Possible objects of study ...................................................................................... 132 5.3 Linguistic worldview in non-native languages ............................................ 141 5.4 Linguistic worldview for Esperanto ................................................................ 145 6 Research problems on the example of a pilot study ................ 154 6.1 The respondents ...................................................................................................... 156 6.2 The questions ........................................................................................................... 162 6.3 Conclusions ................................................................................................................ 180 Final remarks .................................................................................................... 183 The boundaries of artificiality ........................................................................................ 183 A linguistic worldview for artificial languages ......................................................... 184 References .......................................................................................................... 187 Appendix: The questionnaire ...................................................................... 200 Summary in Polish / Streszczenie ............................................................. 211 Summary in Esperanto / Resumo .............................................................. 218 4 List of figures Figure 1 The (simplified) relationship language – culture – perception ............... 19 Figure 2 Relationships between artificial and international languages ................. 34 Figure 3 Federico Gobbo’s coordinate system (2014) .................................................. 74 Figure 4 Classification of invented languages by Albani & Buonarotti (1994) ... 75 Figure 5 The Gnoli triangle (around 1997) ....................................................................... 76 Figure 6 Coloured version of the Gnoli triangle modified by R. A. Brown ............ 76 Figure 7 Jan van Steenbergen’s hexagon (2008) ............................................................. 77 Figure 8 Baron’s natural language spectrum (1994) .................................................... 91 Figure 9 Baron’s computer language spectrum (1994) ............................................... 92 Figure 10 Scale of artificiality/naturalness ....................................................................... 93 Figure 11 Age distribution among the participants .................................................... 157 Figure 12 Native languages of the respondents ........................................................... 158 Figure 13 Usage of Esperanto among the respondents ............................................. 159 Figure 14 Usage of four most popular languages according to activities ........... 160 Figure 15 Usage of Esperanto vs. native languages according to activities ....... 160 Figure 16 In which language would you read a book if the original language was not known to you? .................................................................................................................... 161 Figure 17 Usage of Esperanto vs. native languages according to activities ....... 161 5 List of tables Table 1 Classificatory matrix (simplified with examples) ........................................... 77 Table 2 Hockett’s design features applied to artificial languages ............................ 84 Table 3 Hockett’s design features in formal languages ................................................ 89 Table 4 Languages and their naturalness .......................................................................... 96 Table 5 Languages and their naturalness II ...................................................................... 98 Table 6 A comparison of pidgins and creoles with ethnic languages and IALs 100 Table 7 A comparison of revitalised and minority languages with ethnic languages and IALs .................................................................................................................. 105 Table 8 A comparison of dormant classical languages with ethnic languages and IALs ................................................................................................................................................ 108 Table 9 Types of artificial languages and borderline cases according to usability .......................................................................................................................................................... 132 Table 10 Approaches to artificial languages and some borderline cases ........... 140 Table 11 Samples according to language ........................................................................ 148 Table 12 The features of a “true” Esperantist arranged by aspect ........................ 177 Table 13 The features presupposed in the formula ‘John is an ardent Esperantist but…’ .............................................................................................................................................. 178 Table 14 Symbols of an Esperantist .................................................................................. 179 6 Introduction Research objectives The main objective of the present dissertation is assessing the possibility of studying the concept of the linguistic worldview in artificial languages in the framework of the theory and practice developed in the Ethnolinguistic School of Lublin. Current research on the linguistic worldview is focused on natural ethnic languages and recently comparative studies have been postulated. The present work is the first attempt to show possible applications of the theory to various artificial languages as well as some limitations resulting from such an approach. Artificial languages are not a common subject of today’s linguistics. They are deemed rigid abstract systems suitable for performing the role of interlanguages in translation, serving as a basis for developing programming languages or describing human thinking. Such an understanding of ‘artificial languages’ is, however, very limiting as it does not include universal schemes, international auxiliary languages or artistic languages. The latter types to the mind of linguists are mere curiosities not deserving any serious studies. As the main subject of linguistics, natural languages are assumed. Yet, the author of this dissertation believes that such a binary division into ‘natural’ and ‘artificial’ not only can be disproved but also is detrimental to the development of linguistics as a science trying to uncover the mechanisms lying behind the human linguistic categorisations. Artificial languages are of considerable interest – since they are designed freely by humans, it may mean they obey the same rules as other human languages (the so-called “natural” languages). It has to be remembered that artificial languages constitute a heterogeneous group and that each subgroup is characterised by distinct properties. These properties determine the scope and limitations of studying particular languages. Therefore, the present work aims at revealing the internal diversity of artificial languages and placing them on a scale between the two aforementioned poles. Moreover, as the artificiality is a gradable property so is the possibility of applying the Lublin framework to various artificial systems. 7 Pre-existing research Until now, little importance has been given to artificial languages. They have been the central topic of several historical linguistic works (Bausani 1970; Higley 2007; Knowlson 1975; Large 1985; Salmon 1972; Slaughter 1982). A considerable amount of literature has also been published on artificial languages as the subject of interlinguistics, especially focusing on international auxiliary languages such as Esperanto (particularly Blanke 1985; also the proceedings of the Gesellschaft für Interlinguistik and several hundred publications by i.a. D. Blanke, the most prominent figure in the field, as well as W. Blanke, A. Duličenko, S. Fiedler, F. Gobbo, Ch. G. Kimura, I. Koutny, J. Lindstedt, Haitao Liu, C. Piron, A. Sakaguchi, K. Schubert and H. Tonkin). Several works on artificial languages from a generative perspective have been published by Alan Reed Libert in collaboration with Christo Moskovsky (Libert 2000; Libert 2003; Libert 2004; Libert 2008; Libert & Moskovsky 2011; Moskovsky & Libert 2009). The concept of the linguistic worldview is not a new one, yet a relatively unpopular one. It has its prominent place in Polish ethnolinguistics, where it is extensively discussed; however, it is little known outside Poland. It has been widely investigated not only from a historical perspective (Andrzejewski 1989; Anusiewicz 1999; Underhill 2009; Żuk 2010) but mainly in connection to some philosophical and practical problems by Polish scholars from various research centres (mainly in the Etnolingwistyka journal, the Wrocław series of Język a Kultura and the so-called “Red Series” – a series of proceedings published in Lublin, and recently in Głaz et al. 2013). The initiator of the idea and the de facto founder of the Ethnolinguistic School of Lublin is Jerzy Bartmiński, whose work is of central importance (Bartmiński 2012a; Bartmiński 2012b). Other scholars publishing on the topic include A. Dąbrowska, A. Głaz, R. Grzegorczykowa, S. Niebrzegowska-Bartmińska, A. Pajdzińska, R. Tokarski and too many others to mention in this cursory chapter. The study of the linguistic worldview for artificial languages is a new idea. It has customarily been studied for ethnic languages. The only research article known to the author of this dissertation which discussed the problem of the linguistic worldview for one of the international auxiliary language, i.e. Esperanto, is the article of Koutny (2010). However, to the author’s best knowledge, the application of 8 this concept to a wider range of artificial languages has been scarcely investigated from the theoretical point of view. The present dissertation raises also the problem of the linguistic worldview of multilinguals, which has not been explored until now. Overview The remainder of the present dissertation is organised into five chapters. Chapter 1 discusses the concept of the linguistic worldview and its historical development. The chapter is devoted to the theory and methodology established in the Ethnolinguistic School of Lublin as well as the variety of proposed definitions and some practical consequences of adopting one of them. Crucial concepts such as subject, perspective and facets are introduced. In Chapter 2, artificial languages are presented from a historical and a typological perspective. Their definitions and classifications are examined. The division of languages into natural and artificial is investigated in Chapter 3. A more complex scale is proposed to describe both natural and artificial languages in a common framework. Artificial languages are analysed according to various sets of properties. A class of borderline cases is described. Chapter 4 focuses on Esperanto as a transitional case between an artificially created language and a fully developed natural language with its own speech community. Chapter 5 is an attempt to apply theoretically the paradigm of the Ethnolinguistic School of Lublin to artificial languages and some borderline cases. It outlines possible limitations to such research based on crucial concepts of the framework applied to each type of languages. The chapter presents also a proposition for potential future study of the linguistic worldview in Esperanto. Chapter 6 is a practical application of the earlier propositions to Esperanto. It presents a pilot study conducted in 2015 in the form of a questionnaire and compares it to some of the results presented by Koutny (2010). The dissertation concludes with final remarks. 9 1 Linguistic worldview 1.1 Short history The central concept of this dissertation, namely the linguistic worldview1 featured in the title, has a long history. According to Jerzy Bartmiński (2012a: 22), the initiator and propagator of the idea in Polish linguistics and the founder of the Ethnolinguistic School of Lublin (henceforth ESL), it can be traced as far back as to Aristotle’s topoi (loci communes), that is, common, generally recognised judgements which are part of the argumentation. However, the first obvious reference to dissimilar characteristics of different languages can be found in Martin Luther’s Sendbrif vom Dollmetschen (1530, and even more clearly worded in Tischreden 5, 5521: “ein ittliche sprag hatt ir eigen art”2). In the 16th c., scholars were interested not only in the difficulties of translating the Bible into national languages but also in the newly discovered languages of the Far East and the ways in which they mirrored the reality (see section 2.2.1). Studying languages was for many a way to arrive at truth or the ideal god-made order of things. This is visible in the thought of Leibniz. He claimed that language mirrors the internal structure of intellect, which is common to all men, all being created by the same god. Thus, there is a kind of “universal grammar” or common traits of the mind. However, aspects of reality are differently realised in particular languages because of diverse circumstances and varied experience, which have made man abandon the order of things instituted by god (Święczkowska 1998: 36f., 58f.). It was only in the 18th century that the idea of the influence of cultural and sociological factors on language and thinking gained a strong foothold in philosophy. J. G. Hamann and J. G. Herder saw the language as shaping thought and as a device reflecting the spirit of the nations (Andrzejewski 1989: 163, 174f.; Anusiewicz 1999: 1 Otherwise known as the linguistic picture of the world. For the discussion of the term see Bartmiński (2012a: chap. 7) and Tabakowska (2013). 2 The quote is very often wrongly given as coming from Sendbrif vom Dollmetschen. The passage of Tischreden reads as follows: “Man kan nicht uberal in Hebreo die wort geben, wie sie sthehen, der sprach nach; denn ein ittliche sprag hatt ir eigen art, und ist ein wort so latum, das man im nicht kan genug thun”. 10 263). They opposed the Leibnizian idea of language merely reflecting thought. Hamann went so far as to declare that “reason is language (logos)” (Andrzejewski 1989: 149). The idea of the linguistic worldview (Weltansicht) explicitly appeared for the first time in the works of W. von Humboldt in the first half of the 19th century (see Głaz et al. 2013: 11–24; Underhill 2009; Żuk 2010; cf. Allwood 1983). Humboldt believed that language is “a work of the subject and his a priori spiritual activity” (Andrzejewski 1989: 150). He stressed the importance of individuals, who, although being “cultivated with and within language, they simultaneously cultivated language by leaving their own personal impressions upon it” (Underhill 2009: 122). Language is an ever-going energeia, through its structure shaping thinking and “a nation’s spirit”, i.e. culture. However, the constraints language enforces on thinking can be overcome by creative use. Changes in culture and human activity also influence language and each specific culture is contained in the nation’s vernacular (Andrzejewski 1989: 154). What follows is that different cultures produced different languages and vice versa. The conception presented by Humboldt, namely, that language allows the users to form a worldview (Weltansicht) through its inherent specific structure has come to be mistaken with the conception of Weltanschauung, that is, socially constructed worldviews, which may be different in the same language (e.g. socialist and Christian views of German-speaking people) but the same across languages (e.g. liberal views of English and German speakers). Weltansichten are language- dependent and at the same time, culture-dependent (Underhill 2009: 55ff.) These relativist Humboldtian views were developed in two diverging yet similar ways: by German researchers such as L. Weisgerber and by American anthropologists and linguists (Boas, Sapir, and Whorf). The German Neo-Humboldtians tried to uncover the worldview by studying the division of languages into semantic categories. Weisgerber claimed that the worldview (again Weltansicht rather than Weltanschauung) is a mental construct lying between the world and the community (Zwischenwelt). It is not a reflection of the world but its dynamically changing conceptual interpretation (Andrzejewski 1989: 181f.). H. Gipper went as far as calling language the “key to the world”, as it is a tool 11 that helps conceptualise the perceived reality. For him, however, semantic analysis would not suffice; morphology and syntax were as vital (Anusiewicz 1999). An important contribution of the German researchers was the general term Weltbild, later on translated into Polish as obraz świata. The American anthropologists were occupied with the grammatical distinctions made by languages. As an immigrant, F. Boas believed in differences between languages stemming from their diversely shaped environments. He also saw direct links between language and culture, not claiming, however, that one necessarily conditions the other. E. Sapir and B. Whorf, in turn, postulated that language shapes an individual’s way of thinking by imposing its categories on the perceived reality (Lucy 1992b; Underhill 2009). As Lucy (1992b: 258) writes: “[in Whorf’s approach] individual thought was inferred from the language analysis and empirically verified by reference to related cultural patterns of belief and behavior. This approach allowed Whorf to emphasize the general significance of language patterns for behavior […]”. Comparative approach was for him the basis of the worldview studies. Both groups believed that cultural traces can be found in language and that people can only access reality through language, which categorises things according to its internal structure (Głaz et al. 2013: 11–24; Pajdzińska 2013). In Poland the linguistic worldview (henceforth LWV) has been explored in the Ethnolinguistic School of Lublin (the main representative being J. Bartmiński), Wrocław (e.g. J. Anusiewicz), and Warsaw (R. Grzegorczykowa, axiolinguists with J. Puzynina). The understanding of the worldview as traces of culture in language and not language in culture positions the ESL as stemming from Polish ethnographic tradition and at the same time from the (neo)humboldtism. The contemporary discussion owes much to cognitive linguistics as well. Bartmiński often quotes the accomplishments of A. Wierzbicka on the one hand and Russian linguists such as J. Apresyan and N. Tolstoy on the other (cf. Chlebda 2013). It is clearly visible that the crucial concepts (discussed in more detail in the following section) of the linguistic worldview theory are taken from many sources and thus the theory presents an interesting example of postmodern agreement between different branches of humanities. 12 1.2 Definitions and crucial concepts Humboldt is most known from his relativist approach to languages. Although claiming the capacity for language is the same in all humans, he also stressed the differences between languages being the “organs of the peculiar ways of thinking and feeling of nations3” (Andrzejewski 1989: 162f.). These differences lie primarily in the worldviews. A classical definition of Bartmiński & Tokarski (1986: 72, quoted in Anusiewicz et al. 2000: 28), modified and expanded first in 20064 by Bartmiński, defines the LWV as “[...] a certain set of judgements more or less entrenched in the language, contained in or implied by the meanings of words, which reveals the characteristics and manners of existence of objects from the non-linguistic world”5. According to the latest definition of Jerzy Bartmiński (2012a: 23) the LWV is a language-entrenched interpretation of reality, which can be expressed in the form of judgments about the world, people, things or events. It is an interpretation, not a reflection […]. The interpretation is a result of subjective perception and conceptualization of reality performed by the speakers of a given language; thus, it is clearly subjective and anthropocentric but also intersubjective (social). […] It influences […] the perception and understanding of the social situation by a member of the community. The focal point of this definition is the dynamic character of the worldview contained in language. The LWV is not only an interpretation but also a social and a changing one. This approach follows closely the idea presented by Humboldt, who is “famous for having rejected the idea of language as a product, claiming that it was an activity, a producing” (Underhill 2009: 30). The individual is in the centre of this activity – produces language as a reaction to the perceived world and these perceptions shape language. However, the inherent form of language may also guide the user through the world. The categories of the language are, admittedly, imposed on thinking, but rather as helping patterns than impassable boundaries. 3 All quotations translated by the author. 4 1st ed. of Językowe podstawy obrazu świata; a book translated and published in modified form as Aspects of Cognitive Linguistics in 2009 (1st ed.). 5 Orig. “[…] pewien zespół sądów mniej lub bardziej utrwalonych w języku, zawartych w znaczeniach wyrazów lub przez te znaczenia implikowanych, który orzeka o cechach i sposobach istnienia obiektów świata pozajęzykowego.” 13 These traits (dynamicity, interpreting nature of language) are not always found in other definitions. R. Grzegorczykowa sees the JOS6 as a fixed structure rather than a fluctuating one; “I would like to understand the JOS”, she writes (1999: 41), “as a conceptual structure established (fossilised) in the system of a given language, hence in its grammatical and lexical properties (word meanings and their connectivity), realised, as everything in language, in texts (utterances)”7. However, language does not simply reflect the world but interprets it (Grzegorczykowa 1999: 42, cf. ibid.: 45). W. Pisarek (1978, quoted in Żuk 2010), being the first one in Poland to use the name ‘językowy obraz świata’, writes that the JOS is “reflected in a given national language” (“odbity w danym języku narodowym”; thus the Polish ‘obraz’ should be translated as ‘picture’ in both; ‘picture’ being a fossilised, unchanging structure captured in one precise moment in time and mirrored in language). A short discussion of all the above can be found in Bartmiński (2012a: 24). Bartmiński quotes also a definition by R. Tokarski (2001: 366), who wants the JOS to be a set of regularities (to which Bartmiński opposes, saying that ‘regularity’ is an abstract concept; 2012a: 23): zawartych w kategorialnych związkach gramatycznych (fleksyjnych, słowotwórczych, składniowych) oraz w semantycznych strukturach leksyki, pokazujących swoiste dla danego języka sposoby widzenia poszczególnych składników świata oraz ogólniejsze rozumienie organizacji świata, panujących w nim hierarchii i akceptowanych przez społeczność językową wartości. contained in the categorial grammatical relationships (inflection, word formation, syntax) and the semantic structures of vocabulary, showing language specific ways of seeing the individual components of the world and the broader understanding of the arrangement of the world, the hierarchy in it, and the values accepted by the language community. The JOS can then be found at all language levels. Still, Tokarski acknowledges the primacy of the lexicon in line with other Polish scholars. The definition does not, 6 The acronym JOS [językowy obraz świata; ‘linguistic picture/image/view of the world] is used here to avoid confusion when quoting Polish definitions, where the full name can be translated in several ways into English. Later on in this chapter the LWV acronym is used again. 7 Orig. “Językowy obraz świata chciałabym rozumieć jako strukturą pojęciową utrwaloną (zakrzepłą) w systemie danego języka, a więc w jego właściwościach gramatycznych i leksykalnych (znaczeniach wyrazów i ich łączliwości), realizującą się, jak wszystko w języku, za pomocą tekstów (wypowiedzi).” 14 however, reveal anything more on the nature of the JOS – is it an interpretation or a reflection? Is it dynamic or static? These dilemmas are not present in another two Polish descriptions of the notion. The first one is a proposition of Anusiewicz (1994: 113, quoted in Anusiewicz et al. 2000: 29): określony sposób ujmowania przez język rzeczywistości (zarówno pozajęzykowej, jak i językowej), istniejący w semantycznych, gramatycznych, syntaktycznych i pragmatycznych kategoriach danego języka naturalnego […] to określony sposób odwzorowania świata dany w pojęciowym rozczłonkowaniu zawartym w języku ujmującym ten świat. a certain way of describing the reality (both non-linguistic and linguistic) by language, existing in semantic, grammatical, syntactic and pragmatic categories of a natural language [...] a specific way of mapping the conceptual world contained in the conceptual segmentation in the language describing that world. It is clearly visible that Anusiewicz does not add anything new to the previous definition of Tokarski except for the pragmatic element8. However, he uses the term ‘mapping’, which might suggest that the JOS here is a reflection of the reality, a picture. The second definition of this type is that of J. Maćkiewicz (1999: 8), who considers the JOS to be a part of a bigger, conceptual picture of the world: Obraz świata to […] odbicie doświadczenia poznawczego jakiejś społeczności, […] określony sposób odwzorowania otaczającego świata, […] takie modelowanie rzeczywistości, które umożliwia człowiekowi poruszanie się w niej. The picture of the world is [...] a reflection of the cognitive experience of a community, [...] a certain way of mapping the surrounding world, [...] a modelling of the reality that enables man to navigate in it. This picture can be present in different semiotic systems, one of which is language. Therefore, the JOS is the picture found in a language. It is an abstract model of the reality being a part of the linguistic competence of the user (Maćkiewicz 1999: 10). Clearly, the term ‘picture’ is justified here, as the concept presented is a static9 mirroring of the reality, however dependent on the subject(s), their cognition, 8 Tokarski’s article appeared for the first time in a book under the same title in 1993, published by Wiedza o Kulturze. 9 This word also used by Maćkiewicz herself (1999: 12). 15 experience and their perspective. Interestingly though, she supports both Bartmiński and Grzegorczykowa in claiming that language is an interpreter of the reality (Maćkiewicz 1999: 12) and a reflection of the speaker’s mentality as well (cf. Bartmiński 2012a: 24). An interesting example is provided by Allwood (2004). He is one of few modern Western scholars giving a definition of a concept otherwise not often described. He defines the JOS (Swedish den språkliga världsbilden, a direct counterpart of German sprachliches Weltbild) as a compound of judgments about the world (similar to Bartmiński, who wants to see each notion in a language to be a bundle of traits as described by various subjects): “ett holistiskt system av tankar som används för att identifiera, karakterisera, kategorisera, förklara och förstå många (potentiellt alla) fenomen i omvärlden” (“a holistic system of thoughts used to identify, characterise, classify, explain and understand many (potentially all) phenomena in the world around”). Allwood writes about a system of thoughts, which might remind of the claim Bartmiński makes about the objects of the linguistic worldview (2012a: 67): they are mental objects, not a reflection of the reality (especially if one considers such objects as unicorns or fairies). Thus, it is clear that Bartmiński admits that some objects are created by language. The linguistic worldview not only cannot be a simple mapping of the world (as there are no unicorns) but also cannot create the reality (the existence of the word ‘unicorn’ does not create one). The only entities the language produces are mental ones – either images of the real world objects or of culture (Bartmiński 2012a: 13). This view is supported by Maćkiewicz (1999: 11f.), who states that words are merely naming the elements of culture and not creating the reality. The LWV is thus a reflection of mentality and not of the reality (Bartmiński 2012a: 24). The idea of an intermediate plane (i.e. language) between the world and the individual’s mind noticeably derives through Weisgerber’s Zwischenwelt from Humboldt’s concept of language not merely as a mental structure portraying the reality but as a complement to mind interpreting the nature of objects and of our own thinking in order to clarify ideas (Andrzejewski 1989: 159). R. Grzegorczykowa (1999: 42, 45), however, does not agree with this, claiming that language cannot create the objects of the LWV and that the entities humans talk about are real world objects (even though differently portrayed). 16 A question is then what to do with culturally created notions, such as ‘unicorn’ (on the details of language-culture relationship in the LWV paradigm see section 1.3). Nevertheless, all three scholars believe that language is a tool by which the world is interpreted. The reality as perceived by the individual is shaped in diverse ways across languages and cultures and the choices made in the speech community are then stored in language. Such an operation of shaping is called in the ESL framework ‘profiling’, that is, describing a perceived object in terms of facets (its characteristics). The experiential frame is the base to which contribute “various cultural codes” (Bartmiński 2012a: 89). The interpretation is done by homo loquens and it is the position of the subject and not the position of the object talked about that plays the most important role in the LWV theory of Bartmiński. The very name of the central notion shows that the theory in this variant is subject-related: the view (or vision) must be someone’s view. Not surprisingly, Bartmiński (2012a: 76) recognises the second variant, the object-related one (‘picture of the world’, a direct translation of the German sprachliches Weltbild). Also here the position of the subject and their point of view is important but the focus lies with the static portrait of the entity. Viewpoint as a parameter of perceiving plays a significant role in the theory of the LWV. It is a cultural factor, “a set of directives shaping the content of words”, while the perspective is a bundle of semantic properties of the object resulting from the point of view/viewpoint10 (Bartmiński 2012a: 77f.). A perspective can comprise several different viewpoints and therefore is a more comprehensive notion. The structure of a cognitive definition and the perspective described in it depends largely on the point of view assumed by the subject (for more details see Bartmiński 2012a: 81ff.). Therefore, the subject and their viewpoint can be reconstructed by a careful and detailed profiling of a notion. The profiling is a categorial arrangement of facets, i.e. bundles of judgements made by the subject from a particular viewpoint, which form the cognitive definition. The object in the definition is described in terms of stereotypical, recurring features. 10 ‘Point of view’ and ‘viewpoint’ are treated synonymously by Bartmiński, although the preferred term is the latter one. 17 Bartmiński (2012a: 63–65) understands stereotypes as “stable connections of meanings”, stored in collective memory and reproduced as topoi (utterances purely semantic in form, not fixed), formulae (established combinations transparent semantically to some extent) and idioms (formally fixed combinations, “without clear semantic motivation”). The cognitive definition is a contextual one, that is, one containing connotations of the object described. Such a definition corresponds to ‘a definition through postulates’ of Ajdukiewicz, being “an arrangement of sentences (‘postulates’) that contains the word being defined in various contexts and that meets two conditions: of non-contradiction (i.e. it has a solution) and of non-ambiguity (i.e. it has no more than one solution)” (Bartmiński 2013b). All these basic concepts in the ESL theory clearly indicate that a worldview cannot exist without its most important element – not only a subject but also a community with a common cultural background, without which the facets cannot be reproduced and reconstructed. 1.3 The relation language – culture – thought The problem of the relations mentioned in the title of the section is a complicated one when it comes to the ESL theory. A whole paper by Łozowski (2013) is devoted to the influence of language on culture and vice versa. The linguistic worldview notion would suggest that the core element is the language. Does it, however, influence the culture, or is it influenced by it? Where is the place of cognition and thinking? It is generally agreed that the programme proposed by Bartmiński treats culture as an inseparable part of the worldview and that co-linguistic 11 data (i.e. culturally determined behaviour) are an important material for studies (Maćkiewicz 1999; Żuk 2010). Although Anusiewicz, Dąbrowska and Fleischer (2000) propose that the status of culture be elevated in their theory of cultural worldview, this, in fact, adds nothing new to the LWV programme. The only difference is the suggested material for research: it is based on the standard variety of Polish rather than on folk varieties and includes facial expressions, gestures etc., which the ESL does not in any case exclude. 11 In Bartmiński (2012b) the term is ‘ad-linguistic’. 18 Bartmiński admits that the linguistic worldview can have two interpretations: subject- and object-oriented. The first one (a preferred one in the ESL) is focused on the performance of the speakers, the pragmatics and social considerations. The second one is language-oriented; the object is contained in language itself (Bartmiński 2012a: 76). According to Łozowski the subject-oriented interpretation can be drawn directly from the term ‘linguistic worldview’, which syntagmatically corresponds to ‘[językowy] [obraz świata]’ and suggests that conceptualisations entrenched in language derive from cultural experience (2013: 352). This view might be supported by Allwood’s claim (2004) that a worldview is a system of thoughts and therefore a crucial part of culture. The object-oriented interpretation requires language to influence culture, which is a view coming from the works of Sapir and later Whorf (at least partly). Resolving this issue is not made easier by Bartmiński himself. On the one hand, he states that culture is one of many components of language (2012a: 9), on the other, he claims that culture includes language (2012a: 12). Łozowski (2013: 364–366) rightly points out that Bartmiński, calling his programme ‘cognitive’, suggests the subject-oriented interpretation, whereas writing that “language conditions culture, for without it one cannot participate in culture or in social life” (Bartmiński 2012a: 12), he assumes the views of Sapir and his followers. The very definition of the LWV (a socially intersubjective interpretation of reality entrenched in language) together with the claim that the ESL investigates stereotypes implies the primacy of culture over language. However, continuous references to Sapir (although Bartmiński explicitly rejects the strong version of the relativism hypothesis) could witness to the opposite. Łozowski (2013: 367), quoting Bartmiński, claims the solution to the problem is to acknowledge the bridging position of values. However, even though Bartmiński writes that values lie at the foundation of both language and culture, he does not indicate the source of them other than the human subject. Does it mean that values arise in the mentality/cognition? Or are they produced within culture (which would suggest that culture takes primacy anyway)? Let us look at how other scholars approach this issue. Anusiewicz (1994: 28, quoted in Anusiewicz et al. 2000: 26) believes that observations of the world 19 (the cognitive act) chosen by a community create linguistic categories being an interpretation of the observed world, which in turn influence the cognitive categories. It is visible here that cognition and language are mutually related although the categories are culturally dependent (the social factor is decisive). Grzegorczykowa (1999: 40f.) presents the Ogden-Richards’ triangle – words are connected to real-world objects through mentality of the speakers. Only in the last paragraph of her paper Grzegorczykowa mentions the place of culture, which considerably affects the communication (1999: 46). Maćkiewicz (1999: 11f.) approaches the problem similarly to Anusiewicz: a human subject perceives the objective reality and categorises and interprets it by means of language, which is a reflection of their mentality. The influence of culture is decisive in the process of creating new objects in the worldview. The LWV is a common cognitive basis, to which every member of a community must refer. Therefore, Maćkiewicz adopts also the weak version of the Sapir-Whorf hypothesis by saying that language “facilitates and suggests such and no other interpretations” (1999: 18). It seems that this model of a cycle of influence (see Figure 1) is not only widespread but also accepted by Bartmiński. He argues that both language and culture are “an artifact of human cognitive activity” (2012a: 102) and that cultural patterns should be included in the experiential frame along with conceptualisations (2012a: 89). This not only follows closely the assumptions of Humboldt, who claimed that language helps shape/structure thoughts and at the same time that world-perceiving shapes the language in use, but is also compatible with the conception of Underhill (2009: 134f.), who differentiates between world-perceiving and world-conceiving (the latter one visible in texts): perception categories in language culture human subject objective reality mutual influence Figure 1 The (simplified) relationship language – culture – perception 20 • world-perceiving, for the changing and developing perception we have of the world, • world-conceiving, for the changing and developing manner in which we draw that world into the realm of thought and form concepts and frameworks to represent things and our experience of the world. According to Underhill (2009: 135) also the Weltanschauung might be split into three, this time, different notions: ‘cultural mindset’ (i.e. general, relatively rigid social conception of the world), ‘personal world’ (i.e. the individual’s mindset) and finally ‘perspective’ (i.e. interactive, shifting viewpoint). It is the interactions of the individuals that shape the worldview and therefore the language. It can be concluded that thinking, language and culture are inextricably intertwined in Bartmiński’s approach. This should not be surprising given the fact that the idea of the LWV comes from Humboldt, for whom thinking and language were complimentary activities of human spirit cultivating culture and being cultivated by it (Andrzejewski 1989: 153f.; Underhill 2009: 65f.). 1.4 Methodology Scholars working in the ESL framework propose taking into account different levels of language from vocabulary to grammar (Anusiewicz et al. 2000; Bartmiński 2012a; Grzegorczykowa 1999). Bartmiński (2012a: 71) explicitly expresses his conviction that in order to obtain “content adequacy” (i.e. the inclusion of folk knowledge) in a cognitive definition of an X, the researcher may refer to several different types of data: the language system, texts, questionnaires with native speakers and sociological and ethnographic data (cultural use of an object and speakers’ behaviour in relation to it). However, the cognitive definition may well be based on only one of these types. Only recently, the EUROJOS project postulated that the examination is to be based on a wide array of data to account for a language in its dynamic whole. Such an approach to language as a changing structure is based on the belief of Humboldt who considered it as an activity, an energeia. Any definition of language should therefore be based on individual instances of parole (see Andrzejewski 1989: 157), that is real-life data. In his paper on the role of etymology for the LWV reconstruction Bartmiński (2013a: 235) elaborates on the material and its features: 21 rekonstrukcja [językowego obrazu świata] opiera się na szerokiej bazie materiałowej, odwołuje się do semantycznej analizy kategorii gramatycznych i słownictwa, do łączliwości leksykalnej (stałej i okazjonalnej) i metafor, do struktury logiczno-semantycznej i treści tekstów zarówno kliszowanych (tj. różnych gatunków folkloru: przysłów, zagadek, pieśni, bajek itd.), jak też kreowanych (w ich warstwie presuponowanej, implikowanej). Przydatne dla rekonstrukcji JOS są teksty wywołane, tj. odpowiedzi na pytania o rozumienie słów przez ich użytkowników oraz opisy rytualnych zachowań i wierzeń dotyczących nazywanych przedmiotów. W rekonstruowaniu JOS znaczący udział ma też analiza etymologiczna […]. [the LWV’s] reconstruction is based on a broad material basis; it refers to the semantic analysis of grammatical categories and vocabulary, the lexical connectivity (permanent and occasional) and metaphors, the logical- semantic structure and content of texts both clichéd (i.e. diverse folklore genres: proverbs, riddles, songs, fables etc.) as well as created (in their presupposed, implied layer). Useful for the reconstruction of the LWV are elicited texts, that is, answers to questions about user’s understanding of words and descriptions of ritual behaviours and beliefs about the objects referred to. In reconstructing the LWV the etymological analysis also plays a significant part [...]. Word etymology not only shows how language users conceptualise the world and reveal the choices lying behind the words but also helps the researcher to decide what the internal hierarchy of meanings of a notion should look like and to prioritise primary meanings (Bartmiński 2013a: 236; Underhill 2009: 108). These assumptions have been developed to the fullest in the EUROJOS project (Abramowicz et al. 2009; EUROJOS 2008). The methodological instruction12 distinguishes several sources of material within system and “real-life” data. The data should be extracted from dictionaries (with the reservation that they do not reflect the typical worldview of the everyday language and that only general entries should be considered while specialised meanings left out), texts, corpora and questionnaires. The system data should include the “whole network of lexico-semantical relations” both paradigmatic and syntagmatic:  hypernyms and hyponyms  opposites  synonyms  derivatives (word-formative and semantical) 12 A methodological instruction: the principles and stages of proceeding in the development of the entries within the EUROJOS research programme (available at: http://ispan.waw.pl/default/images/eurojos/eurojos_instrukcja_ma_ibg_wch_jb_26_ii_2011_.doc) 22  complexes and collections  collocations/phrasemes  proverbs (treated as minimal clichéd texts; the proviso is that the original and not the borrowed ones should be primarily studied) Great modern monolingual dictionaries are the source of basic lexeme definitions with their collocations and relevant quotes. Bartmiński (2012a: 67, 71) remarks that the cognitive definition (both the material analysed and the metalanguage) should be primarily occupied with the colloquial variant of a language. However, the boundary between the scientific and the colloquial is blurred. Stylistically neutral usage examples should be extracted from modern texts and corpora (about 200-300 contexts) balanced in respect of the style (popular scientific and journalistic) and political orientation (both left- and right-wing journals; this follows the assertion that Weltansicht is not the same as Weltanschauung; see sections 1.1 and 1.3). However, the researcher has to bear in mind that stereotypical judgements are not introduced as simple assertions but rather hidden in presuppositions and therefore require thorough analysis. It is important to focus on those judgements that are statistically reproduced. Instances of individual judgments can be included provided they are culturally and ethnolinguistically relevant. The last step in the procedure is a questionnaire with only one obligatory question, namely “In your opinion, what is a true X like?” Other than this one, questionnaires can contain open-ended questions as well. These include (cf. Bartmiński 2012a: 132–148, 178–198) supplying only one word best describing the X, naming the objects characteristic of the X, supplying a noun to an adjective, providing synonymous expressions or filling in the blanks (collocation test). The study sample should be a balanced one with at least 100 respondents. Interestingly, Bartmiński (2012a: 132– 148, 178–198) differentiates between ‘ideal’, ‘typical’ and ‘true’ features (‘true’ corresponding to Lakoff’s ‘real’). It seems that ‘ideal’ represents an exemplary X (prescriptive view), ‘typical’ represents an average X (descriptive view), whilst ‘true’ combines both. Therefore, a questionnaire with a question about ‘true’ features of an X is sufficient although the other two types may also be included. Stereotyped judgements about notions may further be studied through Osgood’s semantic differential. Bartmiński (2007) proposes a three-step procedure in which relevant attributes should be selected. Step (1) is the said questionnaire, 23 in which the respondents are prompted to provide one feature of an X. Based on the responses the researcher (2) creates a set of antonymic pairs as a base for (3) the subsequent semantic differential. It is proposed that the antonyms should be placed on a scale graded from 3 to 0 and back to 3 (3 being ‘extremely’, 2 – ‘as usual’, 1 – ‘a little, some’ and 0 – ‘neither A nor B, neutrally’). This approach is well advised. It does not impose on respondents biased, mechanically created antonymic pairs and allows for studying languages non-native for the researcher. The semantic differential is not only useful in assessing the strength of single notions but can be even employed in analyses of synonymous or semantically close notions. Bartmiński (2007: 79f.) shows that the results differ for such synonymous notions as chłop ‘peasant’ and rolnik ‘farmer’. The answers are treated as texts (of a specific genre; see Bartmiński 2012a: 179) and can be divided into denotative associations about the object coming from the user’s extralinguistic knowledge and connotative associations about the object’s name coming from linguistic knowledge. Their relevance should be tested through the ‘but’ test: ready-made sentences should be given to the respondents to assess their acceptability (Bartmiński 2007: 82f.). Alternatively, the respondents could finish such sentences themselves, from which their presuppositions can be extracted (as in Bartmiński 2012a: 134, 180). The answers in the EUROJOS project are later grouped, coded and statistically analysed13. The material is divided into domains (e.g. social aspect, ideological aspect, physical aspect, etc.) in which specific descriptors find their place. For example, if the respondents give names of non-basic colours (e.g. coral, crimson and burgundy), the answers should be grouped under one basic descriptor/keyword (here: red) and allocated to a domain (here: physical aspect). Features within the domains (i.e. facets of a cognitive definition) are to be presented as “minimal diagnostic contexts”, that is in the form of sentences communicating stereotypical judgments of the object. In other words, these are not abstract names of features but sentences or their equivalents: “The horse pulls wagons,” “The horse is a saddle animal,” “The horse is a healthy animal (as a rule),” “The horse can sense a 13 According to Komunikat po międzynarodowej konferencji pt. „Teoria językowego obrazu świata i metody jego rekonstrukcji. Problem eksplikacji wartości” połączonej ze spotkaniem warsztatowym (EUROJOS-VII) (obtained in personal correspondence). 24 person’s death,” etc. These sentences function in a pragmatic-modal frame that one can express as “the speakers think that...” and relate it to “a stereotypical horse.” (Bartmiński 2013b: 170) AN EXAMPLE OF ANALYSIS To better understand this concept an example is in place. M. Grzeszczak in her 2009 paper reconstructs a cognitive definition of DEMOCRACY. She uses three complementary sources of data: system data, questionnaires and texts from two daily gazettes. In the paper a political understanding of democracy is accepted, as a political system in which a specific form of government is practised [out of six dictionary definitions]. On the basis of three types of data, the author identifies the defining (base) features of democracy, correlated with specific aspects of the concepts. (2009: 83) Grzeszczak recognises three senses of the concept: the economic, the social and the ethical one. Her cognitive definition is broken down into facets of both “system” and semantic type. After each facet’s name, a brief explication is presented with an explanation, as follows14: 1. [Name and its etymology] Democracy is a word borrowed from Greek through Latin and it belongs to Europeanisms. 2. [Collections] According to consulted dictionaries, democracy co-occurs with Rzeczpospolita (Pol ‘Republic’). In analysed texts democracy co-occurs with concepts-values, e.g. democracy and justice 3. [Collocations] stable democracy, liberal democracy (in dictionaries), modern democracy, true democracy (in texts) 4. [Synonyms] the rule of people 14 The version presented here is a much-shortened one. 25 5. [Opposites] dictatorship; in texts democracy is opposed to communism 6. [Who is the source/subject of (superior) authority?] nation or people in dictionaries, people or most citizens in questionnaires 7. [What is democracy based on? / What is the foundation of democracy?] According to respondents democracy is most closely linked with ethical values such as freedom/liberty 8. [What does guarantee that democracy works?] the rule of law understood as properly functioning law (questionnaires) and tolerance 9. [What does democracy give to / guarantee to people] freedom and political rights to all citizens (a dictionary), free elections (a dictionary, questionnaires) 10. [Results of democracy / What democracy brings:] optimal conditions for development of every man (questionnaires) 11. [Democracy as an object, goal, and sth desired / What does democracy require?] According to a dictionary, democracy needs fighting for. In light of text data, it needs to be learnt, supported and built. 12. [What is democracy NOT? / What is a contradiction to democracy?] As examples compounds with -kratía are given. 13. [What are the diseases of (modern) democracies?] corruption (in both questionnaires and texts) 14. [What restricts democracy? / What are threats to democracy?] globalisation (texts) Such an approach is presented in and recommended by the Słownik stereotypów i symboli ludowych (‘Dictionary of folk stereotypes and symbols’), the flagship publication of the ESL. 26 The paradigm presented here describes in much detail types of material and ways of collecting it. Bartmiński specifies also how relevant attributes of an X should be chosen through Osgood’s semantic differential and the contradiction test. A shortcoming of this framework is the scant use of corpora. The methodological instruction recommends that only 200-300 contexts should be taken into consideration. Why two hundred and which ones these should be is not indicated. Indeed, corpora may serve as major source of automatically extracted collocations (a more reliable one than even big monolingual dictionaries) and concordances with their frequencies. The fact that the data can be automatically generated significantly shortens the analyses as well as increases their accuracy and reproducibility (Kamasa 2014). Although Bartmiński postulates taking into account all possible connotations of an X (in which a corpus would be helpful, preventing omissions), he also notes that some of them may “turn out to be occasional or coincidental” (2012a: 68). The problem of extraction of the criterial features is still an unsolved one. The closest to a solution comes Bartmiński in his “Kryteria ilościowe w badaniu stereotypów językowych” [‘Quantitative criteria in in studying linguistic stereotypes’; 2007, first published in 1988] where he states that qualitative and quantitative measures should be collated (i.e. the results from contradiction tests and from frequency and/or rank lists). He remarks that semantic acceptability of contradictory statements correlates with a feature’s rank. For example, it is acceptable to say “He’s a doctor but a poor one” because the feature presupposed here (wealth) is a stereotyped high-ranked characteristic of doctors. The correlation is, however, unidirectional (Bartmiński 2007: 83): obejmuje cechy o wysokich wskaźnikach wyborów, natomiast nie pozwala na wykluczenie cech ze środka czy z końca listy; np. cecha postępowy ma przy rolniku zaledwie 21% wyborów, ale zdanie o kimś, że jest rolnikiem, ale nie jest postępowy — jest zdaniem „normalnym” […]. Znaczy to, że o ile wysoka ranga cechy implikuje akceptowalność zdania z presupozycją tej cechy, o tyle niska ranga cechy nie implikuje jego nieakceptowalności. [it] includes high-rank characteristics but does not allow to exclude characteristics from the middle or the end of the list; e.g. the feature progressive of a farmer is at only 21% of choices, but a sentence that someone is a farmer but is not progressive – is a “normal” sentence […]. This means that while a high rank of a feature implies acceptability of a sentence with the presupposed feature, a low rank does not imply its unacceptability. 27 This problem is not easily solved. Even though corpora facilitate collection and preliminary preparation of linguistic material, automatically generated word lists are merely a starting point. The data are subject to subsequent subjective decisions of the researcher. A key problem with this framework, though, is that there are no indications as to how the results should be grouped and processed, how the domains and descriptors are chosen and the facets distinguished. For example, Prorok and Głaz (2013: 187f.) write that [the defining] sentences will be arranged in special semantic categories (facets), that is, groups treated as homogenous from a certain point of view: names, categorization, complexes and collections, oppositions and gradation, origin, transformation, appearance and properties, or actions directed at iron. How these “special semantic categories” are selected does not follow. The authors seem to suggest that the researcher should rely on their native competence or knowledge of competent judges. Niebrzegowska-Bartmińska (2015: 32) describes a facet as wiązka cech, składająca się na eksplikację, odkrywana drogą analizy materiału, a nie narzucana z zewnątrz, a wtórnie traktowana też jako siatka porządkująca materiał i ułatwiająca porównywanie opisów. a bundle of features, constituting the explication, discovered through analysis of the material and not imposed from the outside; secondarily it is treated as a grid ordering the material and facilitating comparison of descriptions. The material is considered here as “revealing its own structure”, an assumption which may be challenged. After all, even if the material shows some patterns, it is the researcher who decides subjectively which domains “reveal themselves” and imposes such and no other grid on the data. It is also very important to note, that facets may serve as the tertium comparationis (TC) in multilingual research. As said, there is some degree of arbitrariness in establishing a full list for a concept. Besides, such a procedure requires native or near-native knowledge of the language in question, making comparative research by one researcher difficult or impossible. The problem of the TC is discussed in detail by Bartmiński (2012a: 214–218), who begins by noting that the choice of an object of comparison can be made from two perspectives: onomasiological and semasiological. Both are problematic: in the onomasiological perspective “the comparative procedure is relatively 28 straightforward in the case of objects unambiguously identifiable” (Bartmiński 2012a: 216), but abstract concepts differ much across languages and cultures; in the semasiological perspective establishing clear boundaries of meanings (division of the world into lexemes) is challenging. He proposes two solutions: working on a concept in a specific cultural sphere where in unclear cases a reference can be made to their common source (e.g. Latin in the case of the Mediterranean culture) or applying a universal semantic metalanguage. Again, both seem insufficiently non-arbitrary. Firstly, there is no guarantee that a concept in all analysed languages comes from the same source. Furthermore, the concept in question has to be first well scrutinised. This means, in fact, that as the TC can serve the concept in any language if only first described in detail. Secondly, the Natural Semantic Metalanguage (NSM), to which Bartmiński references explicitly, also can be criticised as arbitrarily devised or, at least, as being imposed in advance instead of coming from the data at hand, the latter plainly advocated by the ESL15. To conclude, I would like to quote G. Lazard (2001: 365), who declares that the solution to the problem of the TC is to form hypotheses and elaborate sets of concepts for each grammatical domain for use as a tertium comparationis. Such concepts are logically arbitrary and necessarily based on intuition [my emphasis, I.S.]; however, the intuitions behind them are better for being inspired by a large body of experience with linguistic structures. Only via empirical research can these hypothesis [sic] be validated. If they are not validated, they have to be replaced by others that better account for the data. I am afraid there is no other possible procedure for typological comparison. 15 For a discussion of the NSM see Theoretical Linguistics (2004). 29(3). 29 2 Artificial languages 2.1 Definitions A comprehensive definition of an artificial language is hard to find, partly because this branch of linguistics has been neglected or looked down upon (a famous quote by Chomsky goes “Esperanto is not a language. It’s just parasitic on other languages”). In fact, the very name is controversial and many synonyms can be found in the literature. The term in itself comprises a good deal of various, only remotely related codes and systems. It may refer to both langue and langage (see Lyons 1991). Therefore, a clear definition is needed to draw a boundary between what is considered natural and the rest of the systems. Under the name ‘artificial’ the following languages are often placed (cf. Albani & Buonarotti 1994: 9; Carlevaro 1989: 177; Eco 1997: 2f.; Sakaguchi 1998: 26–28):  programming or machine languages (COBOL, Assembler, C#)  formal languages (propositional calculus)  experimental languages: o philosophical (Toki Pona) o logical (Lojban) o pasygraphies (Pictopen) and pasylalies  international auxiliary languages or planned languages (hereafter IALs; Esperanto, Novial)  artistic languages (Klingon, Quenya)  normative languages: o superdialectal (Rumantsch Grishun, Standard Arabic) o standard literary languages (Ausbausprachen) o revived (Cornish) o classical languages (Sanskrit, Latin)  controlled languages (Caterpillar Fundamental English)  reconstructions (Proto-Indo-European)  pidgins and creoles  oneiric languages and glossolalias. 30 Here the term ‘artificial’ is used as a broad name for all languages that are the result of deliberate and conscious (creation and) planning. Therefore, the last type will not be examined in detail because it does not meet the requirement of being deliberately created, although it is discussed in the classification of Albani & Buonarotti (1994) (Figure 4 in this chapter). It is worth mentioning that in their dictionary the term ‘imaginary’ is preferred to ‘artificial’ or ‘invented’ because it stresses the role of fantasy and imagination in the process of creation of fictitious languages. Their understanding of the term is very broad and the creators of the languages described in the book range from professional linguists to outcasts of the society (Albani & Buonarotti 1994: 8): […] con il termine «lingua immaginaria» intendiamo semplicemente una «lingua non naturale», dove l’attributo «naturale» sta ad indicare una lingua il cui apprendimento avviene per trasmissione orale dai genitori e dall’ambiente circostante. Sotto questo profilo «immaginaria» è ogni lingua di tipo artificiale, frutto dell’elaborazione a tavolino di una o più persone […]. […] by the term ‘imaginary’ we mean simply a ‘non-natural language’, where the attribute ‘natural’ denotes a language which is learnt through oral transmission from parents and from the surrounding environment. In this respect, ‘imaginary’ is any type of artificial language, the result of work at the desk of one or more people. A shorter list of systems called ‘artificial’ is given by Blanke (1997: 3): 1. Regularized and standardized literary language, as distinguished from dialects […]. 2. Ethnic languages, highly regularized to maintain them at a particular stage of development (Sanskrit, Church Latin) or to modernize them (Modern Hebrew, Bahasa Indonesia, Landsmål). 3. Consciously created languages to facilitate international communication […], that is, planned languages. 4. Nonredundant, formulaic, or symbolic languages to facilitate scientific thought […]. 5. Programming languages for computers […]. 6. Machine languages for automatic translation. The existing definitions of what an artificial language is are very general and often vague. Fettes (2005) barely enumerates several systems which can be classified as artificial. The entry in Malmkjær (2002) highlights the need of a speech 31 community to call a language natural and the fact that artificial languages are deliberately created: An artificial language is one that has been created for some specific purpose or reason, as opposed to a natural language, such as those spoken by most speech communities around the world, which is normally thought of as having evolved along with its speech community, and for which it is not possible to find some ultimate source of creation. The machine codes and various programming languages we use with computers (see artificial. intelligence) and the languages of logic (see formal logic and modal logic) are all artificial languages […] Likewise, Bausani (1970: 7) states that the artificiality of a language stems from it being created non-spontaneously and non-naturally (“nicht-naturbedingt”). A similar definition is given by the Columbia Encyclopedia online (2014; under ‘international languages’): An artificial language is an idiom that has not developed in a speech community like a natural tongue but has been constructed by human agents from various materials, such as devised signs, elements or modified elements taken from existing natural languages, and invented forms. It does not follow exactly which languages are artificial. A pidgin has also been “constructed by human agents” from diverse components “taken from existing natural languages”, but its naturalness is never questioned (probably because an unconscious character of creation). In the same way, revived languages such as Modern Hebrew or Cornish can be treated as artificial but typically are not (further details in section 3.3). This problem is mentioned by Bartlett (2009). At first, the intention is emphasised: An artificial language is a language that has been deliberately designed for a purpose by one person or a small group of people over a relatively short period of time. (Adapted with permission from a definition by Richard K. Harrison, personal communication, 2004.) Later on, however, the author of the entry briefly mentions the difficulties that arise when this particular definition is used. Should pidgins be considered artificial? Are reduced languages like Basic English natural? The author also indicates that although the definition includes programming languages, these are not discussed, as they do not serve the purpose of human-to-human communication. 32 Both Large (1994) and Kennaway (2010) (even though the entry in the 3rd edition of the Routledge Linguistics Encyclopedia has been revised and seriously modified) offer definitions analogous to the previous ones. Some other names for artificial languages are given, such as ‘conlangs/constructed languages’ and ‘planned languages’. Some (similar) constraints on the categories ‘natural’ and ‘artificial’ are put by Duličenko (2001). He lists three structures: ‘planned languages’ (or ‘international artificial languages’), ‘language constructs’ and ‘ethnic languages’. Planlingvo (PL), aŭ alie internacia artefarita lingvo (IAL), estas objekto, kiu troviĝas inter lingvokonstruo (LK) kaj etna lingvo (EL). LK estas plene produkto de racia kreado ellaborita de unu homo aŭ, plimalofte, de malgranda kolektivo. EL estas spontana-racia kreado ellaborita dum longa tempo de unu konkreta etna kolektivo. LK estas pankronia – nur de momento de ĝia ensociiĝo komencas kreiĝi ĝia diakronio: EL, male, estas elkomence socia kreado kaj tio ĉi estas bazo de ĝia diakronio. LK estas refoja, ĝi kreiĝas laŭ similo al EL; EL estas tiamaniere elkomenca laŭ sia genezo kaj uziĝas kiel bazo por konstruo de PL. Planned language (PL), or in other words international artificial language (IAL), is an object that lies between a language construct (LC) and an ethnic language (EL). A LC is fully a product of rational creation of one person, or, rarely, of a small collective. An EL is spontaneous-rational creation elaborated during a long time of one distinct ethnic collective. A LC is panchronic – only from the moment of its socialisation its diachrony begins. An EL, on the contrary, is from the beginning a social creation, and this is the basis of its diachrony. A LC is secondary; it is created in resemblance to ELs. An EL thus is primary by its origin and is used as a basis for construction of PLs. An interesting case is presented by Gobbo (2012). He discusses the taxonomy given by Lyons (1991), while pointing out some problems connected with it. According to Lyons, there is a four-class division of naturalness (see detailed description in section 3.2.2). Gobbo studies this classification in detail coming to a conclusion that the only types of systems which can be called artificial without doubt are the programming languages (2012: 190). However, Schubert (1989: 9) considers such systems (along with machine and formulaic languages) to be “highly restricted subsets of a language” and does not account for them in his article. All those definitions are in fact alike and leave room for many borderline cases. It would not help to look at definitions of language in general, as there are probably some hundreds of them, producing cases that are even more problematic. Just to give an example: Britannica (2014; under language) quotes three linguists 33 who explicitly write that language is “vocal” (Bloch & Trager) and a combination of “speech-sounds” (Sweet). That would rule out sign languages. Gobbo (2012) requires language to undergo “graphisation” and “socialisation”. While the latter condition is not surprising in light of the definitions given above, the first one seems odd because it would mean denying the name ‘language’ to a large number of small non-codified ethnic languages. Gobbo explains, of course, that he merely means a higher degree of abstractness and not the necessity of an alphabet, while ruling out the possibility that language would be simply equated with parole. In the interlinguistic literature the term ‘artificial’ as opposed to ‘natural’ is regarded as “crudely misleading” (Schubert 1989) because it suggests that languages created to facilitate international communication are in fact identical to machine or formulaic languages. As has already been mentioned, the name ‘artificial’ encompasses various systems and it will be used throughout this dissertation as a general name for all languages that underwent any kind of planning. Other names have also been used throughout history: ‘universal’, ‘international’, ‘auxiliary’, ‘constructed’, ‘planned’, and ‘invented’. Okrent (2010) calls those languages ‘invented’, whereas Bausani uses both ‘artificial’ and ‘invented’ “als Sammel- und Oberbegriff für die Universal- und Geheimsprachen”; (“as collective and generic term for the universal and secret languages” – note that the latter are not cryptic codes but concealed, secretive languages; 1970: 7). The titles of Eco’s (1997) and Large’s (1985) books already hint at which terms the authors prefer: ‘perfect’ (alongside ‘universal’ in the text) and ‘artificial’ respectively. Blanke (1989) distinguishes between two groups of terms that sometimes overlap, i.e. the term ‘constructed/artificial’, which points towards the creation of the language and ‘planned/universal’, which describes the language’s function. This dissertation uses the term ‘artificial’ rather than the currently popular ‘constructed’ to avoid confusion between all artificial language-systems and modern day hobby constructed languages also abbreviated as conlangs. A simple and clear explanation of how to distinguish between ‘universal’ and ‘auxiliary’ languages is given by Liu (2006: 44). For him (and many interlinguists, i.a. Blanke) the term ‘universal’ denotes languages created from the 17th c. onwards meant to be common for all and forever and to replace ethnic languages. The term ‘(international) auxiliary’ and synonymous ‘(international) planned’ is used to 34 represent languages created to facilitate human communication but not necessarily to replace existing languages. This is shown in Figure 2. Figure 2 Relationships between artificial and international languages. Reprinted from Liu 2006: 44 All definitions given above describe the artificial language as a consciously/deliberately invented language with a known creator. In fact, this is a “genetic” trait and does not say anything about the development of artificial languages and the internal variation of the group. If such a binary division is assumed, the only languages not clearly belonging to either natural or artificial are controlled languages (e.g. Basic English), created by deliberately limiting a natural language. Such definitions leave much room for borderline cases, these being controlled languages, revived and revitalised languages (where the historical continuity is broken, the language has no or few native speakers and the revived version is a common effort of a group of enthusiasts, e.g. Modern Hebrew), linguistic reconstructions (hypothetical languages recreated by linguistics), pidgins (simplified vernaculars based on several languages created on purpose but linguistically naïvely) and finally various types of signed languages. A more detailed discussion of these cases taking into account various characteristics other than origin is presented in section 3.3, while the typology of artificial languages will be taken up in section 2.3. 2.2 History and motives of creation Most definitions of artificial languages call attention to the fact that such languages are created for a specific purpose. There are ten motives listed by Bartlett (2009). Some languages are intended to “replace an entire family of languages” (Tutonish), other to be used as auxiliary systems for international communication (Esperanto). There are of course languages designed for artistic use (Quenya, Klingon) or for 35 personal enjoyment (hobby languages on the Internet). Along the secret languages, one might find mystic ones, although some of them are not conscious creations. A few languages belong to the category of philosophical and experimental systems created to test some hypothesis (Loglan, Láadan). Some cater for “special communication needs”, e.g. Blissymbolics for people with disabilities. The last two types are languages “allegedly usable in psychoanalysis” (aUI) and those for communication with extra-terrestrial life forms (Lincos). Bausani (1970: 16–43) gives six types of what he calls “Spracherfindertum” (“language invention”):  linguistic taboo (language engineering)  borrowings from prestigious varieties and creation of compounds  relexification in argots and pidgins  child languages  oneiric languages and glossolalias  poetic creations and riddles A similar, very detailed list of language creation cases is offered by Meyer (1901, quoted in Sakaguchi 1998: 28–31):  spontaneous or intentional disruption of natural development (lexical creations, language games etc.)  spontaneous or intentional changes (children’s secret languages, nicknames etc.)  remodelling of the usual linguistic material (baby talk, jargons, skaldic poetry, purism etc.)  artificial languages in narrow sense (Volapük, Esperanto etc.)  sound-symbolism (glossolalias, riddles etc.)  abstract language creation (a priori languages)  arbitrary creations (taboo languages)  sign systems (sign languages, gestural languages, flag semaphores, symbolic languages and formal languages) Meyer concludes by stating that only the last five examples can be regarded as artificial languages because they are wholly artificial, while the first three only partly. It has to be indicated, though, that sign languages have been proved natural and that 36 this list is unclear in many places (on sign languages see sections 2.2.1 and 3.3.3; see also other classifications by Bausani in sections 2.3.1 and 2.3.2). All of these types can give rise to new artificial constructs; however, they mostly only amount to a small part of the future language. Nevertheless, the main reason for creating a language is probably the striving for perfection of one of many kinds. As Eco (1997: 2f.) notes, it can be perfection in terms of function/structure resulting in a priori systems, in terms of universality resulting in a posteriori languages, and in terms of practicality with the result of polygraphies (i.e. universal scripts). Nevertheless, the motivations go even further, leading to attempts to rediscover the original language of Adam or to reconstruct the primordial language. Historically, Bausani (1970: 15) divides the artificial languages into three categories:  primitive languages and those of the Middle Ages  Renaissance and the 17th c.  auxiliary languages of the 19th and 20th centuries This, obviously, is a very simple division and the timeline of artificial languages is much more complicated. Even the beginnings are very difficult to determine. As one of the oldest examples Albani & Buonarotti (1994: 15) give the language of the frogs of Aristophanes (405 BC). As for a work referring to a possibility of an artificial language, the first one is probably Cratylus by Plato (4th c. BC). Some fragmentary information is available on the language of Galen (2nd c.) and afterwards the gestural language of Beda Venerabilis (7th/8th c.; see Duličenko 1990). However, any sensible timeline or a list of artificial languages is in fact impossible to contrive because there exist hundreds if not thousands of “languages” created for artistic purposes only, in order to make a book or a poem more colourful and interesting and, more recently, myriads of schemes invented “for the sheer joy of it” and published on the internet. The most comprehensive lists have been assembled by Albani & Buonarotti (1994; no less than 1100 pre-internet era languages), Duličenko (1990; exactly 917 IALs from a language project of Galen of the 2nd c. until 1970s) and Okrent (2010; exactly 500 from Lingua Ignota of the 12th c. up to Proto-Central Mountain of 2007). As for the newer languages, the website 37 www.langmaker.com (closed in 2009, although some mirror websites still maintain the copy) was the best source of knowledge with over 1000 entries. Because of the limited scope of this study, the workings of authors “who poetically distort natural language such as Lewis Carrol” (Higley 2007: 9) will not be discussed here. Only two short sections (2.2.4 on artistic languages and 2.2.5 on “modern glossopoeia”) are devoted, respectively, to some major literary creations as well as to the newer schemes found on the Internet. Thus, the languages excluded are those created consciously but “for fun” (“bewusste, fast spielerische”), such as children language games or poetic creations, as well as unconscious glossolalias (“unbewusste Zwangsreden”; both quotations from Bausani 1970: 79). Bausani (1970: 38ff.) argues that most pre-Renaissance artificial languages were mystic or ceremonial attempts of substitutions and additions into ethnic languages (“primitive languages”). He lists five types: taboo, magic, secret, glossolalias and ceremonial ones. Thus, although mentioning two languages before the Renaissance and briefly describing the Renaissance itself, Bausani devotes most attention to languages from the 17th century onwards. Having in mind the purposes of artistic creations (and sometimes lack of material for analysis), many assume that the earliest artificial language is Lingua Ignota from the 12th c. created by St. Hildegard of Bingen (see Bartlett 2009; Large 1985: 3; Okrent 2010: 10). Higley (2007) seems to believe Lingua Ignota to be the first consciously created and systematic invented language. Bausani (1970: 76f.) also claims that the language cannot be classified as a simple glossolalia. As the second more developed artificial language both Bausani (1970: 89f.) and Higley (2007: 66) indicate Bālaibalan from around the turn of the 15th and 16th c. It is worth mentioning that Bausani actually calls Bālaibalan “the first real language invented in an educated milieu” (1970: 83). Those two languages are the only widely known languages that have been consciously created before the 17th century. Other examples from the time include glossolalias, ceremonial formulas and gibberish of mystic works (Bausani 1970: 44– 71; Higley 2007; Large 1985: 3). 38 2.2.1 Universal languages The seventeenth century brought the advent of universal schemes. These were not only spoken languages but also pasygraphies, gestural as well as musical languages. The need arose for various reasons, some very old ones and some newly discovered. The religious conflicts of the 16th c. led to the slow rise of the vernaculars. Later on, local languages grew in importance in printing and in education, which resulted in secondary babelisation of the scholarly world in the second half of the 17th c. (in 1687 Newton published Principia Mathematica in Latin but in 1704 the language of Opticks was English). Latin, which was the sole language of education and the only language taught, experienced its decline and there was no other language to replace it anytime soon (French rose to importance in the last decades of the century). The failure of Latin seems surprising, but at the same time, it can be easily explained. It took too long to learn properly and the results were not of satisfaction: scholars could read and write it but weren’t able to communicate fluently face to face because the differences in pronunciation rendered any conversation almost impossible (Large 1985: 7). A common view was that the language had been somehow “corrupted” through the centuries of using. And not in spite of but because of the many attempts to “resurrect” the classical version, Latin remained dead. As Knowlson (1975: 28) noticed, “[…] the Renaissance insistence upon a return to pure classical Latin meant that this language tended to appear to many to be increasingly unsuited to modern needs”. Interestingly, such problems never occurred in the Islamic world because there was always only one standard Literary Arabic, which made any universal language dispensable (Bausani 1970: 92f.). Nevertheless, there was a need of communication between the East and the West. More and more merchants travelled to the far reaches of Asia. They did not speak Latin, which anyway was of no use to them in China or India (Large 1985: 6–8). The discoveries made in the Far East and in Egypt aroused the interest in pasygraphies, i.e. universal scripts. The belief that one symbol meant exactly one idea or represented exactly one object seemed very attractive. “Analogies were found in musical notes, Arabic numerals, Chinese ideograms and Egyptian hieroglyphics, all of which it was believed provided a representation of concepts which could be 39 understood regardless of language” (Large 1985: 11; see also Slaughter 1982: 85). Many believed quite wrongly that Egyptian hieroglyphics and Chinese characters were not representations of words or phrases but mirrored the reality as it is; the Egyptian symbols bore, however, some resemblance to the objects they signified, whereas Chinese ideograms were a matter of convention (see Eco 1997: 160, 212f.; Knowlson 1975: 15f.; Large 1985: 12f.). The confusion of languages brought to attention the biblical myth of Babel and the beginnings of language itself. The Bible described how the first man named all the things in the world according to what they truly were, which would mean that being an onomatothete ‘name giver’, he was also a nomothete ‘law giver’ (Higley 2007: 6; also Eco 1997: 8). The scholars of the 17th century tried to handle the disarray by attempting to find the original language of Adam. Claims were made that this language was Hebrew (Eco 1997: 80–85), and some even argued that it was some Germanic language, e.g. Flemish as claimed by Goropius Becanus already in 1569 or Swedish by Georg Stiernhielm in 1671; Olaus Rudbeck maintained that Swedish was the source of all Germanic languages in 1675 (Eco 1997: 96–98; Święczkowska 1998: 43–45). The interest in the mystic powers of language realised itself in occult and secret works connecting the kabbalah (especially the art of gematria; stemming from the fact that each of Hebrew letters had its numerical and symbolic value) with the art of cryptography and shorthand (particularly two works of Trithemius: Steganographia written ca. 1499 and Polygraphia from 1508, both published posthumously, respectively 1606 and 1518) as well as with the Egyptian hieroglyphs. At the same time less and less scholars believed it was Hebrew that was the first language in the world and decided to – through comparing living languages – find the true original “language of Adam”. Before the comparative method came into use, such attempts were visible in the works of Leibniz, who tried to compare different languages in order to find to common root of them and, through that, the oldest possible form of language (Święczkowska 1998: 49). THE BEGINNINGS: PASYGRAPHIES AND NUMERICAL LANGUAGES All these reasons brought mass creation of all kinds of schemes. The oldest ones drew on the works of Ramon Lull (13th/14th c.) and Trithemius. Many were 40 inspired by Johannes Becher, who designed one of the best known pasygraphies published in 1661 – Character pro notitia universali linguarum (Bausani 1970: 99f.). He was also an inspiration for his contemporary Athanasius Kircher, who wrote Polygraphia nova et universalis (1663, the title being a clear reference to Trithemius). Both these works, as well as an earlier Universal Character (1657) by Cave Beck utilised the idea of creating a universal system based on Arabic numerals (Bausani 1970: 99f.; Knowlson 1975: 21). Beck compiled a dictionary of Latin words and assigned Arabic numerals to every one of them. He added letters indicating grammatical forms (e.g. p meant the personal pronoun and pf its feminine form) and later syllables for specific numbers, which would make it possible to pronounce the words. However, this would mean that the user would have to remember every entry of the dictionary and the number given to it (Eco 1997: 201). Becher’s system was even more complicated. It employed numbers for words (10,000 entries in the dictionary) as well as for grammatical endings. For fear that the Arabic numbers were not known to everyone he also devised a system of graphical representations for them – a hardly legible one (Eco 1997: 201f.). However, the earliest attempts to devise a pasygraphy of such kind, of which there is not much detail known, are the following four (Knowlson 1975: 44–48):  1627, a prospectus of “escriture universelle” by Jean Douet (the only one of the four to be published);  1628, in a letter from Kinder to Beveridge a scheme of a Monsieur Champagnolle is mentioned; little is known about this scheme – the one presented in the letter is most likely Kinder’s own scheme;  1629, a project of an unknown author criticised in a letter from Descartes to Mersenne. An important addition to the list was made by Francis Lodwick (spellings vary; Lodowyck, Lodowick) in 1647. His pamphlet A Common Writing contained a universal character (a kind of shorthand), where a great number of words could be derived from a limited number of roots (radicals) by using a set of grammatical marks in the form of strokes and hooks. Undoubtedly, this was an original idea and a great contribution to semantics. Moreover, he based his classification not on nouns 41 (as it was usually done at that time), but on verbs (Eco 1997: 260f.). The scheme was never successful, though, as it took English as its base and chose the radicals rather subjectively not really being universal. Memorising the list of radicals and the set of additional symbols seemed to be a challenge. Lodwick was well aware of the weaknesses of his scheme and therefore settled on devising another project, namely The Ground-Work, Or Foundation Laid, (or So Intended) For the Framing of a New Perfect Language: And an Universall or Common Writing (1652). Consequently, the project was a revised and much simpler version of the previous one. Its main topic is lexical derivation; Lodwick proposes that a language should be composed of monosyllabic radicals. Several years later he discussed “the natural order of things”, i.e. possible ways of rational classifications of concepts in his Of an universall reall character (preserved in manuscript). The work again followed the Aristotelian tradition (Salmon 1972: 106). 1652 was also the year of publishing a collection of miscellanea by Sir Thomas Urquhart entitled Ekskubalauron (supposedly Greek for “gold from a dung hill”; Large 1985: 19; Okrent 2010: 28). The collection included a proposal for a universal language, which Urquhart never completed. 1653 brought another creation of Urquhart’s, namely Logopandecteision. However, none of his works included a full description of the languages he praised in his first work. PHILOSOPHICAL LANGUAGES The following years provided a whole collection of other a priori projects: Lingua Philosophica from Ars Signorum by G. Dalgarno (1661), an unnamed project from De Arte Combinatoria by G. W. Leibniz (1666), Panglottie by Comenius (1665-1666) and the Philosophical Language from An essay towards a real character by J. Wilkins (1668). Dalgarno’s scheme was actually a dictionary of allegedly logically organised ideas. He broke down more complicated concepts into basic elements for which he developed a polygraphy (Okrent 2010: 47–49). There were 17 classes of concepts, each marked with a letter (Bausani 1970: 99). Dalgarno reduced both nouns and verbs to a single class of radicals. At first, his idea was to attach grammatical particles to lexical radicals (as was common at that time) but later he decided to eliminate this distinction and derive particles from root radicals. Thus, a verb would 42 consist of a root radical and a modified radical denoting an activity (Cram 1992: 197). Some of the classification schemes for the project were to be prepared by Wilkins. Dalgarno was not happy with the result of Wilkins’s work and decided not to include the tables into his system. Unlike Wilkins, he did not wish to develop a character for every notion possible, but rather a neat and concise system of radicals and a simple grammar allowing for combinatory expressions. This disagreement and the apparent inspiration coming from the works of Comenius pushed Wilkins to write his Real Character in 1668. Comenius, known as a teacher and educator, wrote his work touching this topic, Via lucis, around 1641/1642. In it, he merely outlined the principles of a new universal language – it was to be based on nouns and to mirror the natural order of things. The work circulated in manuscript until published in 1668 (dates differ16). A more detailed proposal called Panglottie was described in the fifth volume of his De rerum humanarum emendatione consultatio catholica (probably 1665-166617). Still the noun was the basis for derivation. However, in the same collection of manuscripts (unfinished and only printed in 1966) can be also found the Novae linguae harmonicae tentamen primum in which Comenius turned his attention rather to verbs than nouns (Salmon 1972: 40, 124). At the same time when Comenius began working on Via lucis, another scholar, the aforementioned John Wilkins, published Mercury, or the Secret and Swift Messenger (1641). The work dealt more with cryptography than with universal language and contained comprehensive discussion of various modes of communication. In one chapter Wilkins considered musical notes, Arabic numerals and Chinese script, among others, as systems on which could contribute to creating one basic language for all humanity consisting of ideas and notions rather than words (Large 1985: 28f.; Salmon 1972: 15). However, Wilkins’s scheme of 1668 seems to be the most known and quoted (and probably the most comprehensive), although he would not have published his project but for the earlier motivations coming from Dalgarno and Comenius. Wilkins 16 Thus, Eco (1997: 215