E is for Esperanto

14 05 2017

teach yourself esperantoTry this thought experiment:

A couple learn an invented language and use it with their child who picks it up naturally. The child eventually meets another person who has the same artificial mother tongue. To what extent will they be able to communicate? That is to say, to what extent will the two linguistic systems be aligned?

Or this one:

Two people, each with different L1s, learn to communicate in a lingua franca for which there are no prescribed rules of suprasegmental phonology, such as rhythm and intonation. Will they be mutually intelligible?

Or this one:

An artificial language has been developed that has its own grammar and vocabulary, but not a codified phraseology, e.g. of collocations, idioms, etc. Will a phraseology develop naturally through use? And to what extent will this cause communication breakdown between speakers of the language who have learned and used it in different settings?

As it happens, these ‘experiments’ are regularly put to the test whenever speakers of artificial languages, such as Esperanto, interact. Designed to be an international lingua franca, Esperanto never quite fulfilled its utopian promise, but (according to Wikipedia) ‘up to two million people worldwide, to varying degrees, speak Esperanto, including about 1,000 to 2,000 native speakers who learned Esperanto from birth.’  This last fact must surely excite researchers of second language acquisition and of sociolinguistics, specifically that aspect of sociolinguistics that deals with generational language change. It’s surprising, therefore, that there is little or no mention of Esperanto in the literature of either SLA or sociolinguistics.

The second generation speakers of Esperanto (I would have thought) would provide interesting data for those who are concerned with how language acquisition emerges, especially in conditions where opportunities for input and output are restricted  – which is often the case, not only for speakers of Esperanto, but also for learners of EFL. And it might provide insights into how languages evolve over time within particular speech communities.

For example, it has been shown (Bergen 2001) that children who grow up speaking Esperanto tend not to use the accusative case. (The accusative case is the marking of nouns and adjectives as objects of the verb. In English, the accusative survives in only a handful of pronouns, e.g. who vs whom). Native speakers of Esperanto also ignore a number of complex tense and aspect distinctions that are marked with affixes.

How does one account for these divergences from ‘proper’ Esperanto (i.e. the language learned by their parents) and the language spoken by second generation Esperanto speakers? Are the differences attributable simply to L1 transfer – given the fact that native Esperanto speakers are invariably bilingual? Or is the ‘nativization’ process determined by general (i.e. not language-specific) learning strategies, such as a tendency to overgeneralize rules or to eliminate redundancy? Or is the failure to adopt features of the target grammar, as prescribed by its grammarians, simply an effect of incomplete learning, due, perhaps, to limited exposure and opportunities for use – what SLA researchers might call the premature stabilization of the interlanguage? Indeed, can we talk about ‘interlanguage’ at all, given that there is no agreed ‘end state’ in the acquisition of Esperanto, i.e. there is no native speaker model that has been codified over generations of users?

Or can second generation Esperanto be explained only by recourse to an innate, language-learning faculty, such as argued by proponents of Universal Grammar (UG)? Could it be that second-generation Esperanto offers evidence of universalizing principles? Which also raises the interesting question as to whether any of the features of Esperanto grammar contravene UG, and, if so, have they been shed in the process of nativization? (Another thought experiment: a language is devised which contravenes UG – e.g. has ‘postpositions’, rather than prepositions (‘the bus on’, not on the bus), but has adjectives before rather than after the noun, i.e. a red bus, not a bus red. It is taught to one generation and then acquired by a second. Would the word order discrepancies resolve themselves? If so, in which direction?)



L.L. Zamenhof (1859-1917), architect of Esperanto

The accusative case, incidentally, has an interesting history in Esperanto: Zamenhof – Esperanto’s designer – believed that the presence of accusative forms of nouns and adjectives would allow a more flexible word order. Thus, with accusative markings, the difference between The dog bit the girl (‘La hundo mordis la knabinon’) and The girl bit the dog (‘La hundon morbis la knabino’) requires no change in word order. But, as early as 1895, there was a heated discussion as to its usefulness. So Zamenhof put it to the vote. The ‘accusativists’ won, triggering a separatist movement within Esperanto, and the formation of a breakaway language called Ido, which abandoned the accusative altogether. As we have seen, nativized Esperanto speakers have tended to follow suit.


Esperanto also offers a suggestive precedent for other (theorized) lingua francas, such as ELF (English as a lingua franca), which have no associated culture and few if any native speakers. Thus, the phonetician, John Wells, an accomplished Esperantist himself, has used the case of Esperanto to argue that speakers of a lingua franca for which there is no codified system of intonation (like Esperanto, like ELF) will simply adopt and adapt the intonation of their L1, with little or no prejudice to intelligibility. This is an argument against the explicit teaching of intonation, especially in the teaching of ELF (see the discussion in I for Intonation). On the other hand, transferring idiomatic expressions from an L1 into a lingua franca (such as Esperanto or ELF) should probably be avoided, since these are unlikely to be transparent to one’s interlocutors – a case against teaching phrasal verbs, for example.

In short, Esperanto, even if not the success its original proponents had envisioned, offers suggestive material for re-imagining the acquisition and teaching of English.


Bergen, B. K. (2001) ‘Nativization processes in L1 Esperanto.’ Journal of Child Language, 28.





13 responses

14 05 2017

Hi Scott
As someone with no knowledge of Esperanto, I’d suggest …

(1) “Are the differences attributable simply to L1 transfer?”

Languages as open dynamic complex systems would naturally involve some transfer as with any other language, but certainly not ‘simply’ L1. That would imply no complexity.

(2) “Or is the failure to adopt features of the target grammar, as prescribed by its grammarians, simply an effect of incomplete learning, due, perhaps, to limited exposure and opportunities for use”

That must be considered a possibility, for the same reason above.

(3) “Which also raises the interesting question as to whether any of the features of Esperanto grammar contravene UG, and, if so, have they been shed in the process of nativization?”

I think I’d agree. If we assume the existence of a UG device, then a language which couldn’t be considered to have evolved naturally, such as Esperanto, may encounter more dramatic changes than a language which has emerged over time in response to cultural/ situational factors, because “the structures of language (have not) emerged from interrelated patterns of experience, social interaction, and cognitive processes.” (Ellis and Larsen Freeman, 2009: 2)


Ellis, N. C., & Larsen-Freeman, D. (2009). Language as a complex adaptive system (Vol. 3). John Wiley & Sons.

14 05 2017
Scott Thornbury

Thanks for the comment, Grant. With regard to your last point, I’m not sure that we have to assume the existence of a UG device, as you put it, to explain the dramatic changes that Esperanto has experienced from one generation to another. The same ‘dramatic’ processes seem to be at work in ‘creolization’ (i.e. when a pidgin becomes ‘nativized’) which can possibly be explained as cognitive ‘streamlining’, as the emergent language system yields to the demands of real-time, socially-situated, and functionally-oriented processing, as when lexical sequences are ‘chunked’ so as to become grammatical morphemes (the way that ‘going to’ became a marker of futurity in English), not because they were predetermined by some innate grammar, but because they were a socially acceptable best fit for a recurrent functional necessity.This may explain the way that nativized Esperanto has shed many extraneous morphemes, such as the accusative case, while ‘inventing’ ways of getting round communicative challenges that were not foreseen by its original designers.

14 05 2017
Justin Willoughby

With regards to Intonation, according to Field (2009), listening out for stressed words in a text is an effective metacogntive, and later cogntive, strategy for improved comprehension. Rost (2011) mentions lexical segmentation strategy, which involves identifying stressed syllables in text to identify the words around those syllables. So, for me it seems that knowledge of intonation and specifically prominence plays an important role in a learner’s ability to process spoken English and hence, worth teaching, if in fact the student is going to be using Engilsh as a franca lingua or has an interest in popular contemporary media like TED talks and ‘The Walking Deadˋ. Mind you, things like dubbing and subtitles in L1 make it more than easy for anyone to enjoy these without having to understand a word of English. So, perhaps the usefulness of an understanding of intonation would be much more immediate in interactive listening.

Listening in the Language Classroom. John Field. 2009
Teaching and Researching Listening. Michael Rost. 2011

14 05 2017
Scott Thornbury

Yes, Justin, it’s indisputable that intonation contributes a great deal to the way that spoken language is processed – but it seems that there are universal (i.e. non-language-specific) principles of intonation that allow speakers of a language that has no codified ‘rules’ of intonation to be mutually intelligible – helped by the fact, perhaps, that speakers of a lingua franca are prepared to make more allowances to their interlocutors than is your average ‘man in the street’ when interacting with ‘Johnny foreigner’.

15 05 2017
Justin Willoughby

So I guess what we are suggesting, along with John Wells (2014) and Jennifer Jenkins (2000), is that it is not worth teaching universal principles of intonation to speakers of a lingua franca. I guess that makes sense as it is something that inevitably carries forward from L1. However, not teaching idioms, like phrasal verbs or metaphors, is going to have an effect on the quality of the discourse and the impact it has on the listener. I think that most languages have idioms and unique sayings that reflect history and culture and really capture what someone, even as a lingua franca, really wants to say. This is a chilling quote from 1984 by George Orwell regarding the development of a new language called ‘Newspeakˋ: ” Every concept that can ever be needed will be expressed by exactly one word, with its meaning rigidly defined and all its subsidiary meanings rubbed out and forgotten”.

16 05 2017
Scott Thornbury

Justin, hi: I think that what Wells and Jenkins are each implying is that yes, there are ‘universal’ features of intonation that can transfer into lingua franca communication (but there may be no harm in teaching those features of English intonation – such as nuclear stress – that aren’t necessarily universal and therefore don’t transfer), and that what I am saying is that the less ‘unilateral idiomaticity (i.e. the use of idiomatic language that is understood by the speaker but not by his or her interlocutors) the better. This does not mean that lingua franca speakers should be discouraged from being ‘creative’ with language, but they need to be aware of the possible consequences in terms of miscommunication. As Luke Prodromou (2008) puts it: ‘As far as idiomaticity concerned, L1-users are playing at home, with rules that they can bend according to need; L2-users are playing away and if they break the rules they may be penalised.’ And he adds, ‘It is possible that the relative absence of certain kinds of idiomaticity in SUEs [successful users of English] is the result not of lack of “competence”, but an intuition, on the level of performance, as to where the limits are, as far as idiomaticity it is concerned… SUEs display an avoidance of language which is prone to pragmatic failure.’ (p 239).

Prodromou, L. (2008) English as a lingua franca: A corpus-based approach. London: Curriculum.

14 05 2017
Heidi A. Karow

It’s helpful to consider research and theory in small bites like this, if you are a busy teacher. Thank you for an interesting post.
I’m wondering if there’s been similar comparisons among different groups using sign language… even the “same” sign language, e.g. ASL.
Incidentally, I have a British cousin who knocked himself out as a teenager to learn some basic sign language before meeting his American aunt who was deaf. Sadly, he discovered she couldn’t recognize his signs.
Imagine a world where everyone (deaf or not) could share one common sign language. Oh how hands would fly at international gatherings.
When I spent time in Germany, I was surprised to learn that even countries geographically closer had developed different sign languages. I met a man whose German sister married a Swede. Both were deaf and both had to overcome a silent language barrier. I think they ended up living in Sweden. So there was that problem too, for one of them… I mean reading text in yet another language.

14 05 2017
Scott Thornbury

Thanks, Heidi, for bringing signing into the discussion. Just as with Esperanto, I suspect that there is a great deal to be learned about language acqusition and language processing from signing – such as how language systems adapt to their local ‘ecological’ niches, and how these adaptations can be explained on purely cognitive grounds without resorting to some ‘magical’ thinking along the lines of a hypothesised ‘universal grammar’. Sadly,though, I don’t know enough about signing to take this argument any further. Clearly, it’s time I did!

15 05 2017

I enjoyed this post. What about Israel that created a new language from old? And I think I read somewhere that speaking a language (intonation) helps to make it familiar and helps memory. I wonder how that would work when you’re not likely to nip out to your local shop and speak Esperanto (although if drunk enough, you might).

15 05 2017
Scott Thornbury

Yes, indeed: the only other study of ‘nativized’ Esperanto that I know (Versteegh 1993) draws a comparison with the way that Ivrit (i.e. modern Hebrew) emerged out of Classical Hebrew as it was adopted by a second generation as their first language – and despite fierce resistance by many Hebrew ‘purists’, who felt that the language was being corrupted, when in fact what was happening was ‘the disappearance of many irregularities and idiosyncratic constructions that were common in the written language’ (p. 457).

Versteegh is intrigued by ‘the attitude of one of the pioneers in the revival of Hebrew, who is said not to have corrected the speech of his children, even when they made mistakes, for fear that they would be discouraged from speaking the language. One suspects that the “mistakes” in question were in many cases structural changes introduced by the children during the process of acquisition. If this attitude was typical of most parents in the early stages of the modern language, it is not surprising that these features got a chance to become permanent in the language’.

He adds: ‘there are obvious parallels between the case of Ivrit and that of Esperanto. The most obvious parallel is that in both cases the parents taught their children an “unnatural” language – that is, a language that was not their own mother tongue – because of ideological reasons.’ (p.548)

Versteegh, K. 1993. ‘Esperanto as a first language: language acquisition with a restricted input’. Linguistics , 31.

18 05 2017

Coincidentally, there was an interesting article about Esperanto in ‘Jacobin’ the other day – it focuses on political aspects of the language: https://www.jacobinmag.com/2017/05/esperanto-world-common-language-zamenhof-internationalism

19 05 2017
Scott Thornbury

Thanks for the link, Philip. It’s interesting that the writer talks about the ‘creolizing’ of Esperanto that occurs when it is ‘nativized’, i.e. acquired by a generation as a mother tongue. In the article I referred to in my post, Bergen discusses to what extent Esperanto qualifies as a creole, i.e. a contact language (or pidgin) that has become the first language of a new generation of speakers, such as Haitian Creole. In the case of Esperanto, the language spoken by the parents differs from a pidgin in that it has a complex (if entirely regular) grammar. Moreover, while pidgins tend to be widely used in the local community, Esperanto ‘natives’ have relatively little contact with other speakers, apart from those in their immediate family. Nevertheless, the linguistic changes that occur as the language morphs from ‘learned’ to ‘acquired’ (to use Krashen’s distinction) do reflect the kinds of processes that occur when pidgins become creoles, which suggests (to Bergen) that first-language Esperanto might best be described as ‘creoloid’. ‘a language with creole-like properties that nevertheless is not a full creole’ (p.591). (Why do I find this so interesting, and what does it suggest about the way English might be changing, the more it is used – and acquired – as a lingua franca?)

8 11 2017

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: