In 1958 Jean Berko devised an experiment to see how young children between the ages of 4 and 7 would respond to nonsense words. (“The Child’s Knowledge of English Morphology”) Kids, you see, say that darndest things, much to the amusement and horror of their parents--and Berko, along with his research assistants, thought it would be fun and educational to say some darn things at kids for a change. This experiment, which linguists affectionately call The Wug Experiment, is legendarily famous, because it revealed a vital clue about how children understand and remember language. Rather than merely being a dictionary-like repository of pre-memorized stock words and phrases, the language of the young child displays a creative and efficiently generalized use and recognition of common patterns.
Let me give you an example to show you what I mean.
Quick! Without thinking, say aloud the following words.
Witch-witches,
Cat-cats,
broom-brooms.
Easy, right? You didn’t even have to think about it--probably you didn’t pay attention to the fact that the plural marker is different from word to word. The difference may not seem very important, but it’s still there. In witches you’ve got an ez-sound; in cats you’ve got an s-sound sound; in brooms, you’ve got a z-sound.
Now-- quick!-- without thinking too hard say these words.:
Wug-wugs
Gutch-gutches
bik-biks
I primed you in the previous paragraph to notice it, but you still didn’t have to think in order to put a z-sound on wug to make wugs; an ez-sound on gutch to make gutches; or an s-sound on bik to make biks.
That is...surprising.
I’d warrant that even if you’d somehow heard these words before, you certainly wouldn’t have heard them used very frequently. Why would that be? Because they are xenowords. Like lycanthropes and vampires, xenowords are monsters whose shapes we dimly recognize as coming from reality. They sound like real words, and look like real words, and even integrate themselves surreptitiously into sentences among real words with a real meaning. Yet they hold no meaning themselves. The surprising thing is that, despite not knowing the meaning of these words, you were nevertheless able to fit each of them into its respective pronunciation pattern. You recognized them as words. So what is going on here?
If these acoustically different sounds nevertheless seemed kinda identical to you, that was by design. The -es in witches and gutches; the -s in cats and biks; the -z in brooms and wugs are all the same morpheme. As you recall from my article on the Grimoire’s Spellspeakers, a morpheme is a unit of meaning smaller than a full word. , It can’t show up on its own. It’s forever tagging along with a content word in one way or another.
In this particular case the morpheme would have a meaning like MORE-THAN-ONE, and it would be realized in English as z, ez, or s, depending on the sound that preceded it. There are famous exceptions. Certain words of Germanic origin express more-than-oneness via a change in one of the words’ vowels, e.g. goose-geese; foot-feet; man-men; woman-women.
As native English-speaker, you, of course, have explicit knowledge of specific terms such as witch, cat, or broom, but you also have a more implicit, practical knowledge that more than one witch is witches with an ez-sound; or that more than one cat is cats with an s-sound.
The difference between explicit and implicit knowledge is instruction. Take knowing how to ride a bike. You can have someone explain to you all day how to balance your weight and keep your momentum going, and none of that will keep you from kissing concrete if you lack confidence and practice. Feeling out the bicycle intuitively, reacting instinctually to the road is going to serve you better than trying to consciously calculate the precise degree of each turn. Knowing how to ride a bike is implicit knowledge. The same couldn’t be said for knowing to steer a ship though, and I’m not talking about a rowboat either. I mean a proper whaling vessel with sails and rigging and anchors and miles of rope everywhere. If you wanted to learn how to lower sails and give chase to Moby Dick, you’d need to get Ishmael to tell you how to tie a knot and chart a course and read the stars. You’d need some explicit knowledge in other words.
Now, children’s knowledge of English morphemes, Berko observed, seems to belong to this implicit kind of knowledge. This is because no one explicitly teaches children morphology. No one sits a child down and is like, “now little Timmy, when you want to pluralize a nominal free form, the underlying form of the standard plural suffix is /s/, and it is realized as [s], [z], or [ez] depending of whether the word ends in a vowel, a sibilant fricative, a voiced consonant, or a voiceless consonant.” Goodness gracious, what a mess. I mean, that’s not even immediately transparent to me as a generalization, and I study this shit. When caretakers teach children about grammar, they use actual words and phrases: e.g. “Look Timmy, that’s a dragon. A Dragon. Now there are two of them, there are two.....” and little Timmy, wide-eyed with wonder would reply “Two dragonzzzz”
Now, of course, the next question is, “Why doesn’t any one teach children morphology explicitly? And the answer to this has to do with what a morpheme is. A morpheme is a functional piece of language, a modifier to a full word, a mini sub-word, if you will. So they never show up alone. Thus it would be silly (and nigh impossible) to try and give isolated examples of morphemes to children. Like, you can point to “dragons” or to “a dragon”, but you can’t point to an “-s”, and be like “look, it’s the plural! Or contrariwise, you can’t point to a bunch of dragons and say just... “-s,” in order to mean “there’s more than one dragon.” (Though, now that I think of it, that would be a badass idea for a conlang. I’m bookmarking that for later!)
So, since morphemes never show up absent content words, children are never exposed to just pure morphemes. So...how do they learn them? That’s the question, isn’t it? The problem that puzzled even Plato. Where does knowledge of something come from if you’re never explicitly taught it? And that is where the Linguistic Xenobestiary comes in.
You see, to answer this question, Berko decided the best course of action was to invent a bunch of xenowords and see if children would recognize them as belonging to a pattern, even though the words were novel to them. It was of vital importance that the words be novel. To quote Berko’s abstract, “If a child knows that the plural of witch is witches, he may simply have memorized the plural form. If, however, he tells us that the plural of * gutch is * gutches, we have evidence that he actually knows, albeit unconsciously, one of those rules which the descriptive linguist, too, would set forth in his grammar.”
That is to say, If the children’s knowledge of morphological patterns is implicit, they should be able to apply the pattern to new forms, since a pattern can be generalized. If however, children’s knowledge of morphology was explicit, viz., pre-memorized dictionary entries gleaned from instruction or imitation, then one would expect them to either reject the xenowords out of hand or just flatly repeat them without fitting them to a pattern.
Now Berko investigated a lot of English patterns with his xenowords, including past-tense formation (to rick-->ricked), progressive formation (to zibb-->zibbing), possessives, and derived adjectives. But I am going to focus on his investigations of more-than-oneness, or “plurality,” if you’re being fancy. So how did the experiment work? The children, who were divided between pre-K and first graders, were shown pictures of various creatures, and were given some conversational prompts to produce the correct forms on their own
Here’s the creature that gave the experiment its name--the wug. And isn’t she just such an adorable little thing? I have a good mind to stat her as a D&D monster. She’ll be a chaotic -good tiny fey who teaches children abandoned in the woods how to speak Sylvan. Not every monster in the dark woods wants a Hansel and Gretel sandwich. But, but, but I was allowing myself to be distracted by my other interests. Let us get back to the matter at hand.
In addition to wug the children also had to find the correct plural form for a number of other xenowords which were:
Lun
Tor
Gutch
Kazh
Niz
Kra
Tass
Heaf
bik
What Berko found was that for words where the adult plural was formed by adding an s- or a z-sound, the children got those right almost all of the time. So they had no trouble at all telling the researchers that more than one wug was wugzz with a “z,” or that more than one bik was biksss, with an s. However, for words where the the adult plural had an ez, like gutches, or nizzes, the children of both age groups had a much harder time finding the correct form. Berko reports that they knew what was being asked of them and understood the task at hand, many children even believing that they were being taught legitimate new English words. They would pause and puzzle over these xenowords, commenting on how difficult they were, before making a best guess. What Berko concluded from these results was that children indeed were capable of recognizing and deploying the English more-than-one pattern, but it was as yet incomplete with respect to an adult knowledge. Specifically, Children knew intuitively how to deploy s and z as plural markers, and slowly were acquiring ez as another part of the pattern. It is also noteworthy that in no cases did children give any very idiosyncratic plural like turning wug into weeg. These facts taken together would suggest that children are receptive only to the most prevalent patterns in their linguistic environment, and will systematically avoid marked exceptions in search of more generalizable patterns. (Yang 2016. “The Price of Linguistic Productivity: How Children Learn to Break the Rules of Language)
Even though Berko was ultimately interested in the nature of Natural Languages, his experiment is of vital importance for conlanging, because it reveals something about what is a possible word. You see, it’s often said (and not entirely wrongly) that a word is an arbitrary linkage between a sound and meaning. This definition goes back to Aristotle at least, and it works well enough to go on with, but as Berko shows, words don’t necessarily have to have an external meaning, don’t have to refer to anything in the real world, to be recognized as words. So clearly, meaning is an indispensable part of a word, but sometimes that meaning doesn’t come from the outside world, but rather from the internal patterns of the language itself. You know that more than one wug are wugzz, even though there is no such thing in reality, and you’ve hardly ever heard that word before. The more-than-one pattern in English lets you conceive even of imaginary or unreal entities in concrete, countable terms. When you’re conlanging, playing with language sounds and fitting them into patterns, you’re essentially inventing meaning from meaninglessness. Or at least opening wide the door to let a monstrous meaning in. Such is the power of the Grimoire. Do you see now, reader, why language is magic?
This article was supposed to go on to talk about Lewis Carrol’s poem The Jabberwock, Paul Eluard’s La Terre est Blue comme une Orange; and Paul Andersson’s “Uncleftish Beholding.” as excellent artistic examples of pushing the rules of language to their limit. But as I typed the last paragraph, I realized that I had quite enough material here for an instructive and fun article, and pushing past that limit would likely strain my readers’ patience. Thus I will continue this line of thought in an upcoming publication. I do hope you enjoyed this one. And stay tuned for the next update.