Whenever I am not practising the languages that I am currently learning, I would usually read up on the inner workings of various languages, or how certain writing systems work. In doing so, particularly the former, I would encounter certain linguistics jargon that might just confuse the average reader. I must admit that when I introduce some languages through my posts on The Language Closet, I would make the occasional mistake to dive headfirst into certain aspects of a language that uses linguistics-heavy terminology, especially so in grammar and syntax. So today, I want to start by dissecting, and hopefully, simplifying certain linguistics terms you might encounter when reading up on a language’s grammar, syntax, or morphology (how a language forms words).

Today’s topic will be the different classifications of languages you will encounter when it comes to morphological typology. Simply put, this is how a language forms words through utilising combinations (or lack thereof) of morphemes. Some languages tend to use fewer morphemes in a single word, while others tend to stack a bunch of them, conveying a meaning of an entire sentence (when translated to English, at least) in a single word.

A morpheme, here, is just a single element that makes up this expression. It is the segment that carries linguistic meaning. To count as a morpheme, it has to carry a meaning (as a word or part thereof), it cannot be divided into any further segments without changing its meaning, and it generally has the same or similar meaning when used in different environments or contexts. So, “cat” is a morpheme, as it carries the meaning of a feline, it cannot be divided into further segments without changing the meaning, and it carries the meaning of a feline when used in different environments. The word “cat” can function alone as its own word, and is referred to as a free morpheme.

But similarly, the “un-” in “unthinkable” is a morpheme, albeit unable to function alone as its own word. It is a morpheme that conveys the meaning of “not”, or negation. Similarly, the “-able” is also a morpheme indicating ability to be done. This inability to function as its own word makes such morphemes bound morphemes. It needs to be bound to other morphemes to form a word. Of course, there are various subcategories of morphemes, which we will focus on in a future post.

Now, equipped with the knowledge of what a morpheme is, it is time to explore the different ways languages form words through combining morphemes (or not). It is better described as a spectrum or continuum of the tendency of languages to form words through fusing multiple morphemes, but a dichotomy might sort languages as analytic or synthetic.

So let us talk about one extremity here — isolating languages. These are languages that tend to have a morpheme-word ratio of close to one, meaning that most words only consist of one morpheme or two. As such, these languages typically lack this thing called inflectional morphology almost entirely. This is when words are modified to convey different grammatical meanings, such as modifying “cat” to “cats”, using the morpheme “-s” to convey the plural of the word “cat”. Without inflectional morphology, these languages are able to keep the morpheme-word ratio extremely low.

One might point towards Mandarin Chinese as a popular example of an isolating language. After all, it lacks inflectional morphology for grammatical categories like tense and number. But it does have a rich variety of compound words, where multiple morphemes, usually two, are combined to convey the meaning of one word. Take, for example, the Mandarin Chinese word for cricket, 蟋蟀 (xī shuài). It is a compound word consisting of two morphemes, but each morpheme alone cannot convey the meaning of anything. But if you want a truer example of and isolating language, consider the other variant of Mandarin Chinese, Classical Chinese, where more words consisting of one morpheme exist.

You might hear such languages being classified under another term called analytic languages. Tracing back to its Ancient Greek origins, the word “analytic” carries a meaning of “loosening up”, or “dividing up”. From this etymology, and the previous mentioning of isolating languages, you might guess that analytic languages are languages that tend to break up multiple concepts into constituent root words, but only using modifying morphemes known as “affixes” rarely. This makes analytic languages have a low morpheme-word ratio.

Although we hear of isolating languages as analytic, the term analytic is often used as a relative term, like, English is relatively analytic compared to, say, Russian. For instance, English does not have inflectional morphology for grammatical gender, and grammatical case. It even uses a separate word to denote the future tense, “will”, although the same cannot be said for the past tense. Inflectional morphology also exists for plural nouns, and person and number when talking about verbs. Additionally, compound words in English may be formed by combining multiple bound or free morphemes into a single word. As an exercise, count how many morphemes make up the word “antidisestablishmentarianism”.

The common features of more analytic languages include a more rigid basic word order. For languages like English, “the cat chases the rat” has a different meaning compared to “the rat chases the cat”. This is not necessarily the case for languages like Russian, where inflectional morphology applies for grammatical case. This way, the word can be modified by morphemes that serve the function of grammatical case to denote the subject and the object, or the agent and the object, and the word order in the sentence does not necessarily need to be changed, although different word orders could give emphasis to different words involved.

Now we move towards another umbrella term, called the synthetic languages. Unlike analytic languages, synthetic languages tend to synthesise multiple concepts into a single word. Syntactic roles like various grammatical functions are designated by adding affixes. Some languages do it more extensively than others, and that would lead us to three primary categories of synthetic languages.

The first of the categories is the fusional languages. These languages tend to use a single morpheme that serves multiple grammatical functions like verb tense, person, and number. You may have encountered such examples in languages like French. For example, in nous voudrions “we would like”, voudrions may be split into the constituents vou + drions, vou– deriving from vouloir (verb to want), and -drions denoting the agreement used for the first person plural pronoun “we”, in the conditional tense. This allows for a variety of inflectional patterns, as multiple grammatical functions are encoded by a single agreeing morpheme.

This is where we would find many well-known Indo-European languages being classified. Modern Standard Arabic also has fusional elements, although how it conjugates to agree with person, grammatical gender, and number is a rather unique feature in the language.

A common feature one would observe is agreement, better known as conjugation. This is where a verb is modified by attaching an affix that encodes information about grammatical mood, tense, voice, aspect, person, grammatical gender, and number, whichever that applies to that particular language. Usually, two or three of these functions are encoded by the affix, which is typically attached to the end of the verb. English has some fusional elements, such as “it eats”, where the “-s” in “eats” is an affix used to conjugate with the verb “eat” to encode the functions of agreement with the third person singular pronoun “it”, and the present tense.

The other feature is declension. Like conjugation, declension is the patterns that apply when nouns and adjectives are modified by an affix that designates grammatical case, number, and grammatical gender. This grammatical case attachment would further specify the role of that modified word in the sentence or clause.

Moving up the analytic-synthetic spectrum, we have the agglutinative languages. Unlike fusional languages, agglutinative languages tend to string multiple morphemes together with the word stem, a process known as agglutination. Each of these suffixes typically represents one type of grammatical function. For example, minulla is the Finnish word that translates literally to “on me”, and is separable into the constituents min(ä) + lla. This -lla affix is the adessive case (alongside -llä), denoting the English equivalent of “on” or “at”. However, multiple stems are generally not incorporated into a single word.

Popular examples associated with being agglutinative include Japanese, Korean, and the Finno-Ugric languages like Finnish and Estonian. Bantu languages like Swahili and Zulu are also generally agglutinative. One pattern to note is the relative regular nature of their grammars. There are just two irregular verbs in Japanese, and a couple of irregular verbs in one Bantu language called Luganda. However, Korean has comparatively more irregular verbs, often classified by the letter in which the verb stem ends. Another interesting feature is that, as these morphemes used to agglutinate with the word stem are regular in nature, it is generally easier to separate a given word into its constituent morphemes to yield an overall interpretable meaning.

Now, on the extreme end of this spectrum, we have the polysynthetic languages. Literally meaning “many putting together”, polysynthetic languages tend to have words with a very high morpheme-word ratio. This allows the potential for such languages to encode the meaning of entire sentences into a single word.

A popular example of a polysynthetic language is Inuktut. For instance, the word takilaangujutit means “you are the tallest”, and can be broken up into the morphemes taki + laa(q) + ngu + jutit, carrying the meanings, in order, “tall”, “superlative affix (-est, or most …)”, “verb to be affix”, and “affix denoting present tense second person singular (singular you)”. Other languages include Greenlandic, where an extensive variety of affixes makes it possible to get a morpheme-word ratio of more than 10, and apart from the root word, no other morpheme can appear alone.

Perhaps the main characteristics you would see is the extensive inflections used to agree with number, tense, person, and other grammatical functions, with some languages using separate markers for each function. Sometimes, the incorporation of other nouns into the word can occur to form things like compound words. Some nouns might also be expressed using verbs, and incorporated into the overall word as a morpheme. For example, clothes might be translated as “the thing that one puts on one’s own body”. All of these strategies used by polysynthetic languages would thus yield very long words, packed with information, and of course, a lot of morphemes.

You might also have heard about the term oligosynthetic. It is a largely theoretical proposition by a familiar name in the linguistics sphere — Whorf. Unlike the other synthetic languages, oligosynthetic languages tend to use a limited number, or relatively few (hence oligo-) morphemes when synthesising concepts into a single word.

Now, we must make it clear that a particular language does not fall 100% into a particular category. There is no purely analytic nor purely synthetic language, and sometimes, the lines defining certain categories of languages can be extremely blurred. For instance, polysynthetic languages may also contain features of agglutinative or fusional languages, particularly in how compound words are formed. Remember “antidisestablishmentarianism”? It is an English word formed using agglutination, but English is more analytic in nature when it comes to grammatical functions.

So this has been an introduction to what is the tip of the morphological iceberg in linguistics. I hope that I have broken down terms as simply as possible, and I do hope you have learned something new when exploring grammars of various languages. I would want to continue this series at some point, and I hope you would stick around for more.