1. Introduction

During the last decades, of particular interest is the theory of Berlin & Kay [5], related to colour categorization. According to this theory, a language can have from 2 to 11 basic colour terms. These eleven basic categories are white, black, red, green, yellow, blue, brown, purple, pink, orange and grey. In addition, the authors [5, P. 2-3] argue that there are strict restrictions on basic colour categories in languages that include fewer than 11 basic colour categories: all languages have terms for white and black; if a language has a third term, then it contains a term for red; if it has four terms, then these are green or yellow; if it has five terms, these are both green and yellow; if a language has a sixth term, it is blue; if it has a seventh, then it is brown; if it has eight terms, then it is purple, pink, orange, grey, or a combination of these.

Besides, Berlin & Kay [5, P. 35-36] mention the possibility a language to have more than eleven basic colour terms. A typical example is Russian, which contains 12 basic colour categories, because of the distinction observed between "sinij" and "goluboj", e.g. between the dark blue and light blue, concerning lightness. [6; 13; 11; 8; 15; 16; 18]. Similar results regarding distinctions of blue appear also in languages related to Russian, such as Ukrainian and Belarusian. According to Safuanova & Korzh [18, c. 78], "sinij" is used in lexical collocations with words such as "sky", "snow", and also in others with a negative connotation (“heavy clouds”, “haze”, “harsh”), while “goluboj” is used with words that have a positive connotation (“affectionate”, “soft”, “tranquil”, “serene”). In a similar way, Turkish has twelve basic colour terms, including "lacrivert" (dark blue) and "mavi" (light blue) [14]. It is noteworthy that in Nepali the term “Akashi” is used to indicate the sky light blue [8]. Moreover, in Arabic the following terms are used for denoting blue: “azrock” (blue), “samawee” (light blue) and “khuhlie” (dark blue) [1]. As far as Greek is concerned, the research of Androulaki et al. [2] is the only one supporting that it has 12 basic colour terms, including γalбzjo (light blue). Nevertheless, except blй ("blue") and γalбzjo (light blue), they recognize the presence of γalanу, which they consider as identical with γalбzjo, without discerning any difference between these two terms [2, P. 6, 10). Research about the basic colour terms in Modern Greek has been also carried out by Serakioti & Markopoulos [19] and Serakioti [20].

  The aim of the present research is the experimental examination of the following issues: a) if in Modern Greek there is any difference between the terms blй, γalбzjo, γalanу, b) if the above colour terms can be characterised as basic, fulfilling Berlin & Kay’s criteria [5, P. 6-7].

2. Methodology

In this study, the Munsell colour array, used also by Berlin & Kay [5], was given  (Figure 1) to the subjects of the research, who were asked to find the typical version of blй, γalбzjo and γalanу. More particularly, the second stage of data collection of Berlin & Kay’s experiment was followed [5, P. 5]. The Munsell colour array contains a set of 330 colour chips, of which 320 represent 40 different hues, each divided into 8 different lightness levels (Munsell value). The purpose of this experiment is to check whether there is deviation in the perception of the three colour terms blue, γalбzjo, γalanу, taking into account that Androulaki et al. [2, P. 6, 10] consider γalбzjo as identical to γalanу.

It is worth noting that the survey participants were asked to mention their age and gender, and these data were later processed. To avoiding the effect of the order in which colours were presented, a different order of colours for each participant was followed.

For registering, grouping and statistically processing data, tables were created, which include the serial number of the speaker, the gender, the colour and the corresponding values for its placement on the vertical and the horizontal axis of the colour spectrum (see Appendix). To facilitate data processing and numerical calculations, for the vertical placement in the colour spectrum numbers were used instead of letters (A = 0, B = 1, C = 2, D = 3, E =4, F = 5, G = 6, H = 7, I = 8). Furthermore, on each axis the average and the standard deviation were calculated for denoting the median of each colour, in order to investigate whether all values are focused on a particular point or there is great dispersion around a central value. In addition, the confidence interval was calculated on the vertical and horizontal placement to indicate the maximum error probability (0.05). The outliers were removed, e.g. values exceeding x’+2´s, and the average and standard deviation were recalculated.

The survey involved a total of 60 Greek native speakers, 30 males and 30 females, aged 20-50 years, who had no achromatopsia or dyschromatopsia problems (see 10; 7, 21; 17].

3. FindingsandDiscussion

The results of the present research indicate that the terms γalбzjo and γalanу do not correspond to identical values in the spectrum, while there is significant deviation between them and blй. More particularly, γalбzjo appears to be a darker hue of blй in relation to γalanу, with average (26.7, 3.2), while γalanу appears as a lighter hue of blue in relation to γalбzjo, with average (26.82, 2.79), as it can be seen in Figure 2. Thus, in the horizontal axis there is no difference between the two colour terms, contrary to the vertical axis. It seems that this difference regarding the colour lightness is not accidental, since the average of each point is outside the confidence interval of the other.

Moreover, it should be also mentioned that between these two adjectives (γalбzjos-γalanуs) there are differences regarding the collocations and each of them refers to a particular class of objects. For example, γalбzjos is mainly used in collocations with words such as θбlasa (sea), lнmni (lake), aktн (shore), while γalanуs with words such as uranуs (sky) and also simйa (flag), indicating the blue and white colour of the Greek flag. Thus, it is expected for one to say e.g. γalбzja  θбlasa/ lнmni/ akti, instead of γalanн θбlasa/ lнmni/ aktн or phrases, such as “to γalбzjo tis θбlasas/ tu uranъ” (the blue of the sea/ sky)  instead of “to γalanу tis θбlasas/ tu uranъ”. From the SEK (Corpus of Greek Texts) (Goutsos 2013) this difference is confirmed, since the term γalбzjos is used in plenty of collocations, much more than γalanуs. These are some examples: “to γalбzjo tu pelбgus” (the blue of the sea), “to aperado γalбzjo tou Egйu” (the endless blue of the Aegean Sea), “to γalбzjo tis θбlasas ke tu uranъ” (the blue of the sea and the sky), “γalбzjo akroʝбli” (blue shore), “γalбzja lнmni” (blue lake), and also metonymically “to nisн pu zу нne γalбzjo” (the island in which I live is blue), instead of “to nisн pu zу йxi γalбzja θбlasa” (the island in which I live has blue sea). The term γalanуs appears, mainly, in literacy in collocations such as “γalanуs uranуs” (blue sky), “γalanб nerб” (blue water), “γalanб mбtзa” (blue eyes) and also metonymically, e.g. “to γalanу  tis vlйma” (her blue look).  

Figure 2. Average ofblй, γalбzjo, γalanу in Modern Greek

Moreover, there is a difference between γalбzjos and γalanуs regarding their compounds and derivatives. Thus, in compounding with the prefix kata- the term kataγбlanos (very blue) is used instead of *kataγбlazos. According to SEK, the derived word kataγбlanos appears in collocations such as “kataγбlanos uranуs”, “kataγбlani θбlasa”, “kataγбlana mбtзa”, excluding the use of *kataγбlazos (“*kataγбlazos uranуs”,  “*kataγбlazi θбlasa”,  “*kataγбlaza mбtзa”. Regarding the derivation of these colour terms with the suffix –opуs (-ish), γalazopуs is used instead of *γalanopуs.

 Furthermore, in compounding the terms γalazoprбsino (blue-green) (γalбzjo + prбsino) and prasinoγбlazo (prбsino + γalбzjo) are preferred and not *γalanoprбsino (γalanу + prбsino) and *prasinogбlano (prбsino + γalanу). Also, the term γαlαnуlefkos (blue and white) (γalanу + lefkу) is used, but not the term *γαlazoleukos (γalбzjo + lefkу).

In the research of Androulaki et al. [2] γalбzjo is presented as “lighter blue” and considered as a basic colour term of Greek, on the basis of the frequency of its appearance, while it is also claimed that the term γalбzjos is identical with the term γalanуs [2 P. 6, 10-11]. Thus, the term γalбzjos and γalanуs are both considered as light blue, without any differentiation between them. Moreover, it is argued that there is no difference regarding the collocations of the two terms, since they are both used to denote the colour of the sea, sky and the greek flag, and also the blue colour of the eyes [2, P. 6].

Regarding the methodology of their research, there is considerable inconsistency in the used sample, since the number of the subjects varies in each stage of the experiment (In different stages of the research the subjects were 8 bilingual speakers of Greek and English (6 females and 2 males), 6 Greek speakers from Crete (3 males and 3 females) speaking a particular Greek dialect of Crete, 18 Greek native speakers (10 males and 8 females) and 12 Greek native speakers (7 females and 5 males)), as well as the ratio of the representatives of the two genders (male – female), while only a limited range of ages is covered (19 to 31 years old). Additionally, the selection of the subjects from Crete, who speak a particular dialect of Greek, is rather debatable (as Androulaki et al. [2, c. 18] mention, the six subjects of the research were born and lived in an isolated village of Crete with about 150 inhabitants).

Concerning the presented experiment, we notice a difference between γalбzjo and γalanу regarding the hue. Thus, we can consider the two terms as a variation of blй, given the fact that γalanу corresponds to a lighter hue of blй in relation to γalбzjo, and γalбzjo to a darker hue of blй in relation to γalanу (Figure 4).

For the question if γalбzjo and γalanу are basic colour terms, it would be useful to examine Berlin & Kay’s [5, P. 6-7] criteria for the characterisation of a colour term as basic. A colour term is considered as basic when: a) its meaning does not derive from the meaning of its parts (e.g. “blue” and not “bluish”, “olive green”), b) it is not included in another colour category (e.g. “scarlet” is included in “red” and is a kind of it), c) it is not limited to a category of objects (e.g. “blond” is mainly used with “hair”, “complexion” and “furniture”), and d) it is widely known and accepted by all the speakers (e.g. “yellow” instead of “saffron”).

Based on the above criteria, the terms γalбzjo and γalanу cannot be considered as basic colour terms in Modern Greek, since they violate the criteria (b) and (c). On the one hand they violate the criterion (b), since they are included in the basic colour category of blй and are a subcategory of it, and on the other hand they violate the criterion (c), since each of them appears only in particular collocations.

4. Conclusion

According to the present experimental research, the term γalбzjo is a darker hue of blй in relation to γalanу, while γalanу is a lighter blй than γalбzjo, since there is a deviation of values regarding their lightness on the vertical axis. Moreover, these terms cannot be considered as basic colour terms, since, on the one hand, they are included in a basic colour category being a subcategory of it and, on the other hand, they occur only in particular collocations, violating the corresponding Berlin & Kay’s [5] criteria. Nevertheless, this issue needs further investigation and interdisciplinary approach.

Acknowledgments: I would like to thank Professor Stavros Skopeteas for his support and valuable comments.

APPENDIX. The data for blй, γalбzjo, γalanу in Modern Greek