During the 1970s, there was a flurry of research into what was termed ‘acquisition order‘ or ‘morpheme order‘ studies. (FOR EXAMPLE, Brown 1973, Dulay and Burt 1974 and 1975, Bailey, Madden and Krashen 1974, Fathman 1975, Andersen 1976, Kessler and Idar 1977, Krashen 1977, Fabris 1978, Christison 1979, Makino 1980). These attempted to see if people learned second languages in a predictable order, regardless of the second language being learned. In other words, a natural order of acquisition. The studies also wanted to find out if this order was different to the order of acquisition in a first language.
This research provided strong evidence to suggest that there is an acquisition order, an order that is only slightly influenced by a learner’s L1. Broadly speaking, this order is:
- articles: the, a
- copula be (e.g. she is tall)
- progressive -ing
- plural -s
- auxiliary be (e.g. she is singing)
- regular past tense -ed
- irregular past tense
- possessive -‘s
- third person singular -s
This led some researchers to make a fairly bold claim: that the existence of a natural order meant that instruction had little or no effect on acquisition.
However, when we examine these studies more closely, the contents don’t really match up to the packaging. These studies call themselves acquisition order studies, so how do they measure acquisition? How do they work out when a morpheme or structure has been acquired? You would think they would use some kind of a longitudinal study but in fact the studies I have looked at seem to measure, not, acquisition, but accuracy. (Dulay and Burt 1974 had originally intended to do a year long longitudinal study, but abandoned this in favour of a snapshot).
Accuracy Orders, not Acquisition Orders
Some studies examine learner speech or writing, and measure how accurately learners produce a range of language features, usually morphemes. From this data, they then assume that the more accurate a language feature is, the earlier it was acquired.
But perhaps the most accurate forms are simply the easiest to learn, because they are the most noticeable or because they are the most frequent. In other words, these studies are simply measuring learnability.
Also, mastery of grammatical items may be a lot more complex than acquisition orders assume. Mastery may not be as linear as we might think. Many learners of English don’t acquire the irregular past tense of eat as in:
but through a U-shaped development:
The ‘eated’ and ‘ated’ forms may be the result of a learner modifying their interlanguage to accommodate new knowledge (that the simple past tense ends in -ed). Other grammatical items (e.g. negatives, interrogatives, relative clauses, past tense) may show a similarly non-linear developmental sequence. If you study a learner’s accuracy when they are at the bottom of the U, they will get a low accuracy score even though they are on the way to ‘getting it’; a learner studied at the start of the U will display acquisition, even though they may be about to plunge down to inaccuracy!
It’s also a bit much to lump all irregular verbs into one lump and then say that irregular verbs are acquired later (7th on the list of 9): many irregular verbs are very common (e.g. went, ate, ran etc) and are likely to be learned very early on (subject to U-shaped development); some irregular verbs (e.g. put, bought) are likely to be learned much later.
Also, these studies tend to ignore the idea of interlanguage as a learner’s developing internal language system. If we measure ‘accuracy’ we are simply comparing a learner’s current language system with the target language, something which Vivian Cook has called the ‘comparative fallacy’.
However, what the acquisition (or accuracy) order studies to indicate is that a lot of learner development is systematic and, to an extent, predictable. But that’s about it. Claims from these studies that instruction has no effect are ridiculous: while instruction may not alter the order of accuracy/acquisition, it can speed it up.
Vivian Cook (1999), Going beyond the native speaker in language teaching. TESOL Quarterly, 33, 185-209.
Heidi Dulay and Marina Burt (1974), Natural Sequences in Child Second Language Acquisition