So, it’s finally happening. A couple of years of language enthusiast status and a good year and a half of German learning by immersion, I am taking on one of Europe’s bonafide Behemoths: I’m going to tell you about Hungarian. The Hungarian language is notorious for being incredibly hard to learn and has features not found in many other European languages. This article will take you through some of these features and also raise an interesting question: is the concept of an ‘agglutinative’ language absolute?
The first thing we need to get out of the way is the bit about difference. Hungarian is rather unlike other European languages because it is actually not an Indo-European language. It belongs to the Ugric branch of the Uralic language family and is actually a distant cousin of Finnish and Estonian (both of these belong to the Finnic branch of the family). The Magyars (what the Hungarians call themselves) were originally a nomadic tribe from, you guessed it, the Ural Mountains region and settled in the Carpathian Basin in around the 9th century.
Also, whilst Hungarian shares many similarities as far as structure is concerned, the former two are not mutually intelligible to it. It’s sort of like a Brit listening to a Dutch person, the form of the language is similar, but they don’t understand what is being said. Got it? Good, now let’s move on to some features.
A large factor in what makes Hungarian so difficult to learn is the grammatical devices it employs. Firstly, the use of personal suffixes to indicate the verb “to be” is present in Hungarian: an example of this is the phrase “I don’t speak Hungarian” Nem beszélek magyarul. Hungarian is pro-drop, so the 1st person singular “I” Én, does not need to be present. So, the verb beszél (to speak) needs a suffix denoting the 1st person singular én, one can then read this as if it were a pronoun to determine who we are talking about.
Thus, a word-for-word translation would be “No Hungarian speak-am”. Once you get used to reading the verb “to be” in suffix form as the pronoun, it becomes easier to identify the subject. You also need to get accustomed to reading suffixes all the time, as we shall see later. This can take a while if your native language does not utilise this system, but you’ll get there with practice.
But of course, that is not it. This is Hungarian and if you thought it was going to be that simple, you are in for some disappointment! Each pronoun’s “to be” suffix can take not one, not two but six different forms.
Why is this? Well, Hungarian has two categories of conjugation: definite and indefinite. Definite is used when the verb can take a direct object, like the verb ‘to eat’ in ‘I eat the apple’ for example. On the other hand, indefinite is used if the verb cannot take a direct object in the given context, for example “they were running”. The latter is the variant I used in the previous paragraph.
To add to this, Hungarian features a concept called Vowel Harmony. This means that the vowel you use in the suffix is dictated by its ‘harmony’ with the vowel in the verb stem. There are two main types of stem vowels: front vowels and back vowels. Additionally, there is a third category for stems which contain the vowels ö/ő and ü/ű. Here’s a helpful table to (hopefully) alleviate head scratching:
Pronoun suffix (definite then indefinite) | Back vowels: a, á, o, ó, u, ú | Front vowels: e, é, i, í, ö, ő, ü, ű | ö,ő/ü,ű |
1st person | om/ok | em/ek | öm |
2nd person | od/sz – or ‘L’ with ‘sz’ verb endings | ed | öd |
3rd person | ja/none | i/none | jak -ik/none |
You’ll notice that I haven’t used any plural suffixes here. I want to keep it simple so as not to hit you with too much technical stuff, that way we can focus on exploring the question I posed earlier regarding agglutination. Where does this come from? The ‘agglutination’ hypothesis was proposed by linguist Martin Haspelmath and challenges the idea of morphology categorisation. He rightly points out that this is a fairly recent phenomenon formulated in the early 1800’s, mainly by a handful of German linguists.
The need to categorise languages increased in the wake of the explosion of nationalism in Europe that began around this time; it ultimately caused the concept of a ‘national’ language to become popular and grow into what we know today. But not everyone is convinced that this doctrine holds up today.
Okay, with that whistle stop tour of Hungarian grammar over with, why don’t we discuss a bit more of the Agglutination Hypothesis and see what it’s all about?
As I outlined previously, the idea of categorising languages was drawn up in the early 19th century. It was then popularised in the 1850’s – 1870’s and mainly focuses on three groups: isolate (one morpheme per word), agglutination (many morphemes stuck together) and fusional (where the divisions between morphemes are obscured). You may ask what a ‘morpheme’ is. It’s simply the smallest possible semantic ‘piece’ of information a language offers.
Some examples in English would be -ing, -ly, -ed and -s. As these suffixes need to be attached to a stem to function, English cannot be considered a purely isolate (or analytical) language and is instead deemed a fusional language. This is because it incorporates both affix/suffix morphemes but cannot ‘stick together’ multiple morphemes to express complex concepts. But it does stand that a fusional language can use inflection to convey pieces of information. Thus, the fundamental premise of the ‘hypothesis’ is that agglutination is a concept that can and should be tested, not a hard and fast rule.
Before he runs the test, Haspelmath first predicts that any language which shows agglutination in one area will invariably show it in others. Upon raising this point, he immediately excludes an aspect of language from his analysis: affix alteration. Why? Well, think of the number of affixes we encounter all the time warmer (intensification) Große (adjective inflected for feminine form), les œuvres (Plural form).
This kind of alteration exists in Hungarian too; we have already seen it in fact with the vowel harmony and definite/indefinite conjugations from earlier and can be considered fusional, not agglutinative. Throughout his analysis, there are three distinct characteristic groupings of which 30 languages from a diverse range of families are compared. Interestingly, when the ‘average’ marks of agglutination results are revealed, Hungarian ranks 25th out of 30th. In addition, he marks the language with an asterisk which is he terms languages labelled agglutinative “in literature”. So why such a low score then? Let’s investigate a bit.
Taking the ‘most agglutinating’ language that has been ‘marked out’ as such by the asterisk (Turkish), we can do a little side by comparison. So, let’s take a simple sentence like ‘I am going to the shop”, in Hungarian we have this:
Elmegyek a boltba – “going-am the shop-to”
And in Turkish: markete gidyorum – “shop (market) going-am”
You will notice here to both languages have similar characteristics. Firstly, they are pro-drop languages, although Turkish less so. Both languages feature vowel harmony; the personal suffix um in the Turkish is used to harmonise with the vowel ‘o’ in its stem noun. Without digressing too much, the rules for Turkish vowel harmony are a little different. The vowel “o” in the back position can be classed as either strong or round and suffix vowels only ever harmonise with the ‘back’ vowel.
But our sample sentence reveals one small difference, Hungarian features a definite article ‘the’ with a and Turkish does not. But a rather minute factor is not the strongest indicator of a divergence in the amount of agglutination. For example, the negative behaves in a different way in Turkish, as seen in this sentence “I am not a doctor”
Hungarian: Nem vagyok orvos – Not am doctor
Turkish: Ben doctor değilum – I doctor not-am
We can see two differences here, the first being something I alluded to earlier. Turkish is a pro-drop language, but it seems that when it comes to simple sentences such as these, the personal pronoun is in fact used. The second difference is that the negative in Turkish requires the personal suffix, whereas Hungarian employs it as a standalone morpheme. By definition, this makes the Turkish version of this sentence more agglutinative. Still not convinced? Take a look at a longer sentence like “I am not going to the city”
Hungarian: Nem megyek a varósba – not going-am the city-to
Turkish: Şehre gitmiyorum – to-city am-not-going
In this sentence, Turkish is clearly more agglutinative. This is down to the fact that Turkish verbs have a negative conjugation – gitmiyorum – compared to gidiyorum that we saw earlier. You will notice a change in the Turkish. The 1st person Ben has been dropped, whereas Hungarian’s one-word negation is never changing. Also, from the literal translation break down, we can see that Turkish displays a near-perfect definition of agglutination: little pieces of information ‘glued’ together. Again, we can see clear agglutination in both languages, but more in Turkish.
After this interesting, albeit brief, analysis of the two languages, maybe it’s to make this judgement: a language might not be fully ‘agglutinating’ but that shouldn’t stop us from referring to it as such. Surely, you’d agree with me that the fact that a language can display agglutination is the part that matters.
A drumkit is known to be a ‘loud’ instrument because it has the ability to be and often is. But that doesn’t mean that it has to be all the time, right? In the same way, these two languages certainly agglutinate much of the time, but there are instances where they do not, and share features common amongst fusional languages (whilst both share features amongst analytical languages).
The idea that fusional language be characterised as such is just because they very rarely do something akin to ‘agglutination’, and, when they do, it is normally a small detail like a personal suffix. Languages such as English and Spanish can simply do no more. We will not see a preposition or connective attached to a verb in either of these languages and thus they cannot be classed as Agglutinating. The languages we just looked at can do more agglutination and therefore it makes sense that they be called ‘agglutinating’ languages.
There is something to take away from this though. Due to the centuries-long presence of the Ottoman Empire in South-Eastern Europe, Western scholars became well acquainted with the Turkish language (although less ‘Turkic’ than today’s Turkish). Hapselmath points out that this led them to characterise languages such as Hungarian as “Turkish-like” due to their shared, capacity for agglutination, irrespective of how frequently it may occur. Also, it’s not a mere coincidence that these two languages share concepts such as vowel harmony; another ‘asterisked’ language Swahili also uses this.
So, it’s important that we don’t paint languages with the same brush, but also recognise the similarities they share and the ability to ‘loosely’ categorise. That’s why I think that it’s a good idea to preserve the term ‘agglutination’ to demonstrate a function both of these languages can perform, whilst at the same time not expecting the same level of performance from both. In closing: yes, Hungarian is an agglutinating language, but it should be seen as a feature of the language rather than an overarching form.
Get more interesting language and linguistics content in our magazine. You can subscribe below