First, I just want to apologize briefly for taking a while to update. Sometimes I have loads of free time, and other times, grad school hits you like a freight train and says “HERE, DO ALL OF THIS” (which was, at last tally: re-structure and revise a paper, outline all of the ethical issues my dissertation may have and how I plan to approach those issues, teach English three times a week, do French practice on Duolingo every day, practice Chinese at least a few times a week, do readings for class, work out….and update this blog). I say “freight train” lovingly, though. Grad school is like a really awesome freight train of knowledge and adventure.
But my daily life in grad school is not the focus of this entry. The last time I updated this blog (in the Pleistocene Era, it seems), I talked about how people acquire their first language (henceforth known as L1). I mentioned in that entry that while we can definitely use L1 acquisition to understand how people learn a second language (henceforth known as L2), there are some big differences between the two. One of the biggest is that when you learn your L1, you’re starting from the bottom with naught but some evolutionary aptitude for sponging up language very quickly. When you learn your L2, you’re going in with an entire linguistic framework hardwired into your brain. You’ve learned how to conjugate verbs, how (and when) to make certain sounds, how to put a sentence together, and what you call the objects and actions that surround you. You’re going into your L2 with some fairly strong ideas of how to string words together and express yourself. Under these circumstances, L1 influence (or transfer) on your L2 is inevitable. But is that the whole story?
The history of the study of Cross-Linguistic Influence is quite long (by Second Language Acquisition standards, anyway). It has its origins in the 1950s. The first studies of Cross-Linguistic Influence fell under what was called the Contrastive Analysis Hypothesis (CAH) — this hypothesis stated that people learn through imitation and habit formation, L1 being a very prominent and pervasive “habit”. The difficulties people had in their L2 learning stemmed from the differences between their L2 and their L1, or the linguistic habits formed by their L1. L2 acquisition, ideally, would replace L1 habits with new L2 habits, which could be achieved through (wait for it…) Contrastive Analysis.
In Contrastive Analysis, a researcher compares two languages to determine potential errors and to see what might be easy for learners due to similarities. Essentially, CAH predicts that it will be easy to learn structures in L2 that are similar to structures in L1, and difficult to learn structures in L2 that are different from structures in L1. Ease and difficulty are on a spectrum of sorts, which I’ve illustrated here, using English as the example L1 and Spanish as the example L2:
- Split (Hardest to learn according to CAH) – In Spanish, there are two words that correspond to the English “to be”, and have different functions (ser for more permanent states, estar for impermanent states)
- New Concept – Spanish, unlike English, has gendered adjectives and nouns that must agree with each other.
- Absence – Spanish does not use “do” (its equivalent “hacer”) as an auxiliary or modal verb. Instead of “does not (verb)”, a Spanish speaker would say “no (verb)”.
- Coalescence – In English, third-person possessive pronouns are gendered (his and her). In Spanish, only one possessive pronoun (su) is used for both genders (plus the polite second person form, Usted).
- Correspondence (Easiest to learn according to CAH) – The present progressive forms are very similar in Spanish and English. For example, the English “I am listening” corresponds with the Spanish “Estoy escuchando“.
To some extent, CAH makes a lot of sense. However, there are some major issues with it. Firstly, as many language learners (including yours truly) can attest, sometimes splits and new concepts can be easier to learn, since you don’t need to un-learn something else from your L1. Secondly, researchers have found that many of the errors people make when learning an L2 can be very similar and in a particular developmental sequence, regardless of the structure of their L1. This means that the errors CAH predicts either don’t happen, or would happen no matter what your L1 happens to be. Other times, the errors CAH predicted just didn’t occur.
From the ashes of poor semi-discredited CAH came a new theory, Error Analysis (EA). Error Analysis sought to build off of the problems with CAH and create a new method for improving L2 acquisition. Error Analysis decided to dispense with L1 entirely and just look at the errors the learner made in the target language. A researcher could then identify, classify, and quantify errors, analyze this data, and figure out why the learner was making these mistakes.
This sounds awesome, right? Well, like CAH, EA had some bugs. One of the biggest was that it relied (relies) very heavily on a researcher or educator’s judgment. Unfortunately, most linguists are not telepathic (unless my SLA professor is hiding a rather impressive secret), so it’s easy for a researcher or educator to inaccurately ascribe the cause of an error, or to inaccurately determine the source of an error. However, the step of taking L1 out of the equation was an important step, which leads us to the next development in the study of Cross-Linguistic Influence.
Interlanguage, which was first developed by the linguist Selinker in 1972, hypothesizes that there is an intermediate system between L1 and L2 (the eponymous Interlanguage) that helps a learner synthesize and develop language knowledge, and leads to certain types of L1 influence. Interlanguage is a constantly developing system; some aspects of your interlanguage won’t develop until you’re fairly advanced in your L2 and learn certain structures.
The idea of Interlanguage is arguably one of the foundational theories of modern Second Language Acquisition. It is also an important and fascinating link between L1 acquisition and L2 acquisition; we’re going to discuss this more in-depth in a little bit, but as I mentioned briefly earlier in my section on CAH, a lot of the errors L2 learners make don’t come from L1 or from the surrounding language environment, but are quite similar to errors that children learning their L1 make (which, as you might recall, don’t usually come from what they hear around them). Some researchers believe (and I’m inclined to agree) that the errors language acquirers make are cognitive solutions and stop-gap measures until they can fully understand or figure out new language structures.
The Interlanguage “system”, we can say, is a bit like a laboratory in an L2 learner’s brain where they can make errors in L2 and work them out, just like when they learned L1. Except — let’s keep up the laboratory method here — when the L2 learner was just learning L1, there was nothing in the laboratory yet. Now with L1 under their belt, the L2 learner’s Interlanguage laboratory is a bit more equipped, but the processes and errors, researchers say, are very similar to the ones exhibited in L1 acquisition, no matter what L1 was getting acquired.
So now that we have that nice meaty theoretical foundation, let’s talk about some examples of Cross-Linguistic Influence and errors. Still with me here? Good.
Lourdes Ortega (2009) discusses two huge sources of Cross-Linguistic Influence: Interlingual Identification, and perception of linguistic differences or distance. Interlingual Identification is when learners encounter difficulties due to (contrary to what CAH would have you think) real or perceived similarities between L1 and L2. Sometimes, this means finding something in L2 that lines up with a structure you know in L1. For example, Klee and Ocampo (1995) found that speakers of Quechua (an indigenous language spoken in the Andes by the descendants of the Incas) who spoke Spanish as their L2 would often use a certain verb tense in Spanish to indicate evidentiality (how true a statement is). In Quechua, this is included in the verb. In Spanish, speakers indicate evidentiality with modifying words and phrases, which is what we do in English (e.g. words like supposedly, allegedly, evidently, or seemed to). Learners will use similar or analogous structures in L2 to compensate for differences from L1. But for that transfer to happen, the learner has to perceive that those structures might line up.
Another way learners will try to compensate for different (or overly similar) structures in their L2 is by simply avoiding the really structures they just can’t get yet. If we go back to the idea of Interlanguage and the hypothesis that the errors we make in our Interlanguage are just stop-gaps until we’re at the right developmental stage for structures to be second nature, avoidance makes a lot of sense. If you can express what you mean without using a structure that your brain just hasn’t quite gotten the hang of yet, you’re in business.
Here’s an example of that from my own language learning experience: in Mandarin Chinese, there’s a specific grammatical form for expressing the passive voice (object was verbed by subject). I still have trouble with it now, even after studying Mandarin for several years. In my everyday speech and writing, I almost never use it unless prompted to by a teacher. I should definitely try to use it more, since it’s very common in formal writing, but in my Interlanguage system, I’ve compensated by just using the active voice to express the same thing (perhaps it makes what I say sound more emphatic or less formal, though). On the flip side, some of my Mandarin-speaking friends have said that they find the passive voice in English difficult, and just use the active voice instead to avoid it.
Another error is overuse. Sometimes, when a language learner acquires a new structure in their L2, they use that structure too much, or use one aspect of that structure too much. For instance, Finnish encodes prepositions in suffixes attached to nouns — you wouldn’t say “in the house”, you’d say “house-in” (or something along those lines; Finnish is super hard, guys). When Finns learn English or another language that uses prepositions, they either don’t use prepositions enough (what CAH would predict), or over-extend the meaning of one particular preposition (like “in”) (Ortega, 2009). I’d argue that overuse is a good thing; going back to the idea of your Interlanguage as a language laboratory in your brain, when you’re overusing or over-extending a new concept from L2, you’re experimenting with it, figuring out how it works and trying to apply it to lots of different sentences and situations.
I’ve mostly focused on what’s called “negative transfer”, or errors that arise in L2 learning (either due to Interlanguage development or due to structural differences between L2 and L1), but Cross-Linguistic Influence isn’t always a bad thing. As I mentioned before, sometimes big differences between two languages can be beneficial, since you don’t need to un-learn an existing concept you know from L1. Other times, close similarities between L1 and L2 can be a good thing, since you can use what you know from L1 to understand a concept in L2, or find words that relate to each other. I know that sounds a bit contradictory, but it all depends on what structures you’re learning and how you’re learning them.
In addition, multilingualism, even if the languages you already know are not related to the language you are learning, can help you become more proficient in your target language. According to Ortega (2009), in a study of Iranian students learning English, students who spoke Persian and Armenian, or Persian and Turkish, did better on English tests than students who spoke only Persian and English. So if you want to get better at learning languages…learn more languages.
So what can we get from all of this? In short, your L1 an L2 are definitely going to collide and influence each other (yes, L2 can influence L1), but a lot of L2 development follows a path similar to the one you followed as you developed your L1. Your L1 just gives you more tools to help (or hinder) that development.
Up next: “chunks” in language learning. You’ll see what I mean. Cryptic? I’m not being cryptic.
Ortega, L. (2009). Understanding Second Language Acquisition. Hodder Arnold.