In all efforts put into understanding LUs and their occurrence, a methodology of categorizing languages according to type was proposed, leading to a new branch of linguistic study known as typology in LUs.
Typology as language classification allows for comparison of formal modes among languages, i.e., the internal attributes of languages and the presence of such attribute would establish its place in the respective category
Typology works in a manner of efficiently sorting out linguistic features that are highly noticeable across languages and picking them to be the head of a type. Following which, grouping languages according to these common features would allow linguists to draw patterns and systematically limit languages accordingly (Greenberg, 1969).
Hence, three major areas in linguistics (Morphology, Syntax and Phonology) will be explored and within each area, occurrence of either type of typological universals will be used as plausible explanations for the manifestation and mechanisms behind language universals. In turn, these evidences would be used to postulate for the evolution of human languages
At face value, morphological difference is one of the most glaring differences among languages. Without deeper investigation, language universals can be easily dismissed as void. Every language seems to possess a combinatorial system for morphemes that is not exactly the same as the other language – especially one that comes from another language family. This combinatorial principle, a hallmark of human language, leads to the complex nature of morphological processes and structure. Additionally, the complexity of each morphological structure varies and this is mirrored by the difference which spans from grammatical organisation of morphemes to how they are stored as lexical entries (Evans & Levinson, 2009). Proponents of language universals postulate that language differences can be analysed through a more abstract investigation which would render these differences to be negligible. However, critics are adamant about leaving these differences as they are in light of language diversity (Evans & Levinson, 2009).
Canonical language types
At word level, the difference in how a word is formed can be broadly categorised into three: isolating, agglutinating, and fusional. For isolating language type, there appears to be one-to-one agreement between the words and morphemes. The languages which come close to the isolating type is Vietnamese and Chinese. Meanwhile, in agglutinating language, a word is usually made up of more than one morphemes. These morphemes are distinctive and have consistent forms. A close example would be Turkish. For fusional language, there is no well-defined boundaries between morphemes. A word consists of unsegmented morphemes of different grammatical and verbal categories such as Russian.
However, it would be erroneous to simplistically categorise languages of the world into these language types. Majority of the world’s languages do not conform specifically to one of these. Instead, they “fall between the two extremes on each of the indices of synthesis and fusion.” (Comrie,1989) Index of synthesis refers to the number of morphemes per word while index of fusion refers to number of grammatical categories per morpheme.
In every scientific research including one that relates to language, there is an intrinsic need to search for general laws in the construction of languages. Researchers would then look into extending these laws to its widest possible scope.(Van Der Hulst,2008 ) These laws come in the form of universals where they are claimed to exist at the “deeper levels of analysis and theorising” (Van Der Hulst, 2008). However, the extent to which universals can be applied remain unanswered. Among the substantive universals that are claimed to be applicable to all languages include Verb Affixes (Pinker & Bloom,1990). This claim is immediately counteract by the fact that it does not hold true for all languages. For instance, Mandarin and Malay languages do not mark tense and many spoken languages such as German lack aspect.
In defence of Language Universals, Tallerman (2009) postulates that “examining languages more closely, or at a higher level of abstraction often reveals critical similarities which superficial descriptions can obscure.”
This is highlighted in the example of the Wakashan languages, Nuuchahnult. (Tallerman, 2009) In this language, nominal and verbal roots are not easily distinguishable by morphology. In addition, lexical roots including nouns can take verbal inflectional morphology thus superficially postulating the absence of noun and verb distinction. A deeper level of analysis however, reveals that there is a behavioural differences between inflections on noun and verb . For instance, proper names cannot take on third singular indicative verbal inflection -maa as illustrated below:
1. mamuuk-maa quuʔas-ʔi
“The man is working.”
2. quuʔas-maa mamuuk- ʔi
“The working one is a man”
(“He is Jack.”)
In another part of inflection, the suffix –(m)it applies on both nouns and verb. However, this past tense marker in Nuuchanult conveys the exact meaning of “former” for proper names (4) but it simply indicates past tense when inflected on verbal predicate (5).
4. mamuuk-(m)it -(m)aħ
“I was working.”
5. ʔuunuu ʔani ʔuumiik-(m)it-qa
because that whaler-PAST-SUBORDINATE
“because he was a former whaler”
Nuuchanult language highlights Tallerman’s argument that language universals may not be evident at face value. It takes careful investigation to draw out similar features between languages. (Evans & Levinson, 2009) claim that not all language distinguish their major lexical categories; nouns, verbs, adjectives. But the above examples suggest a flaw in their claims. Perhaps other languages that are seemingly invariant in lexical categories may reveal these distinctions through a careful morphological analysis.
Pinker & Bloom (1990), proposed that there are substantive universals common to all languages. These substantive universals are the building blocks of grammar that would be present in languages as either an explicit inventory or as a consequence of more abstract mechanisms.
Major phrasal categories for example, would start with a major lexical item. That means that if a verb phrase is present, it is because the head for the phrase is a verb. In other words, major lexical phrasal categories are assumed to be present in all languages. It appears to be a sound argument as many English language speakers would agree that the aforementioned categories can be applied to English. In addition, many second language learners are taught the new language first by assuming these categories. Myles (2004) supports this as in her study, it was shown that learners of second language do project the lexical categories in the initial stage and subsequently move on to phrasal categories of varying difficulty level.
Moreover, until recent years, many still agreed with this notion that such categories are universal with Chung (2012) study on Chamorro as one such example. Chamorro is an Austronesian language which has been previously studied and conclusions were drawn saying that Chamorro is a language with unusual lexical category system. These conclusions were drawn based on true observations of the language. However, Chung (2012) argues that though it was based on true observations, it was too small a sample size of Chamorro morphological and syntax to make any “robust conclusions” (pp.50). Upon a further study on Chung’s part, it was found that Chamorro does have the usual categories such as nouns and verbs with multiple syntactic functions. Through this, with compelling evidence of such categories to be found in many languages, the idea of language universals is very much appealing.
However, word classes as examined by Evans & Levinson (2009), seem not to be as universal as one thought. They found counter evidence in varying languages studied by others to refute the claim of ‘all languages having the same word classes’. Enfield (2004) and Hengeveld (1992) both found that not all languages have adjectives or an adverb class. Languages that lack in a certain ‘word class’ would seemingly make up for it through other means like another form of modification with morphology for example (Hengeveld, 1992). Furthermore there are word classes not commonly found in Indo-European languages (Evans & Levinson, 2009).
Ideophones and positionals to name a few, are word classes that integrate syntactically in the language. Ideophones are most typically words that depict an idea of a sound, sight, smell or certain sensory perceptions (Nuckolls, 1996). Japanese is one such language that uses ideophones syntactically in the language. An example below extracted from Torantani (2007) is presented below where the ideophone is embedded in the syntax as an adverb (pp.322).
In the example, gósi-gósi is an ideophone which depicts the sound of scrubbing or rubbing something. It functions as an adverb in the example as it describes the rubbing of the floor. This shows that the ideophone is part of the sentence syntactically.
Positionals can be easily found in Mayan Languages and mainly describe a person or object’s position and form (Ameka & Levinson, 2007; Bohnemeyer & Brown, 2007). On the other hand, classifiers are also a word class that cannot be put across all languages to be universally defined as they are not “identical in nature” (Evans & Levinson, 2009, p.435). Numeral classifiers in East Asian and Mesoamerican Language can classify objects in various ways. It can be classify counted objects according to shape, size, texture, and many more to name a few. As Haspelmath (2007), argues that word classes are not exactly the same between any two languages in morphosyntactic or semantic properties.
Based on the above discussion, it seems that universal traits that could be found in word classes of syntax are not as substantial as one might have thought. Nonetheless, it does seem that with regard to the word classes of the different language it points towards the adaptationist view in the respect of language evolution.
A study of Greenberg word order universals by Shopen (2007), found that across a wide sample languages both related and unrelated showed two sets of word order relation. The first is where dependent word order relations have a correlation with verb and object order. The second showed that word orders are independent of this. Dunn, Greenhill, Levinson & Gray (2011), however, argued that few word order features are correlated. They also argued that the functional dependencies observed between these traits are more lineage specific than universal. That is to say most languages studied in detail and where the apparent universal traits are found are mainly from languages in the same language family rather than across language families. Dunn et al. (2007) argued that within a language family, similar traits are found across these languages while languages from a different family have different traits. This was most likely the product of cultural evolution thus determining the linguistic features thereby forming the current language system (Dunn et al., 2007). The differing view on the existence of word order universal is truly interesting indeed. Cultural evolution that comes into play may prove to be an interesting perspective in the course of trying to find out more about the universals in language with regard to word order in syntax.
Goldin-Meadow, So, Özyürek & Mylander (2008) showed that speakers of different languages with different set of word orders seem to express events nonverbally in the same manner. Participants came from different linguistic backgrounds with native languages that have different word order. It was expected that the event was to be expressed through gestures following the word order of the participants own native language. Results showed that even with different linguistic backgrounds, participants had the same “word order” in expressing the same event through the same series of actions. Interestingly, none of the participants displayed the word order of their language. This study might be a clue that shows an evolution of gestural language as it seem that it is a natural system that humans have when expressing things nonverbally with no knowledge of sign language.
The word order in languages do not seem to be very controversial as little evidence is presented to support the presence of universal traits. However, in the case of the non-verbal communication in terms of the series of gestures and actions, there seem to be a specific way of doing it. This could open a new perspective to the argument of language evolution being innate. After all, with no prior knowledge of sign language, a group of people with no common language between them is able to communicate effectively the desired sentence.
Typological differences in phonology are observed in marked features and phonological hierarchies (Greenberg, 1969).
Marked features refer to the additional articulatory features that are more complex in nature and less preferred in sound sequences.
Phonological hierarchies refer to the preference order of phonology in single phonemes and phonemic clustering across various positions in a word.
For instance, according to Greenberg (1969), nasal vowels require an “additional resonance chamber, the nasal, besides the oral chamber which functions alone in the oral vowels”(Greenberg, 1969, p.476). Nasal vowels are considered to be a marked feature in languages that have both nasal and oral vowels. Similarly, nasal vowels are less preferred over oral vowels because of their restricted acoustic environments, where they can occur in selected word formations. Thus, nasal vowels are placed on a lower position than oral vowels within the phonological hierarchy.
According to sonority sequencing universals (where sonority is loudness of a speech sound), language users would generally prefer consonant sequences (C1C2, where C denotes consonant) with high sonority difference over those with low sonority difference between the sequences.
blif > bnif > bdif > lbif
In particular, Berent, Lennertz, Jun, Moreno, Smolensky (2008) had done a research study on word-initial consonant cluster preference. The guiding principle of the study is that speakers, when presented with perceived mispronounced words, tend to repair them within their perception to be correctly pronounced. For example, speakers of languages with word-initial consonant clusters (i.e. English speakers and the word ‘black’) would be able to perceive blif as monosyllabic while words beginning with lbif would be misperceived as lebif, changing the syllable count to disyllabic.
In the study, they recruited native Korean speakers. Due to their lack in consonant clusters in the beginning of words in their language, it presented a clear contrast with English speakers who are regularly exposed to word-initial consonant clusters. The study posited that if Korean speakers demonstrated similar preference order and syllabic-count judgement as English speakers, their perception of initial consonant clusters would be attributed to reasons beyond exposure to certain speech sounds.
They found that Korean speakers did show preference for consonant sequences in an order that is similar to their English speaking counterparts. Hence this was in support of the hypothesis of the study: “adult human brain possess knowledge of universal properties of linguistic structures absent from their language” (Berent, et al, 2008, p.5324). This instinct could be attributed to the human brain’s capacity in universal linguistic knowledge, advocating for human’s innate ability to produce language.
In Evans and Levinson (2009), they talked about physical characteristics of vocal tract being “the clearest evidence for biological basis for language” (p.433). At the same time, they also brought up another form in which human natural language are realised: natural sign languages.
Sign languages are gestural rather than verbal, and it is created in by people, for people, who are unable to acquire verbal languages, such as Hearing Impaired.
Perhaps, in using sign languages to disprove phonological universals would seem inadequate – after all, signed languages have no need of speech sounds. However, Evans & Levinson (2009) did emphasized that the existence of sign languages account for the “modality-plastic nature of our language capacity” (p. 433). This implies that human brains, to a certain extent, are adapted to be able to create new communication systems where the existing system is not longer suitable for their usage. This further emphasize for the notion of languages are adapted in relation to our cognitive capabilities.
Therefore, the conflicting evidences as illustrated above had drawn an indistinct picture of how phonological universals in languages are supported in Linguistics and served to show how capricious languages are.