If someone were to say
to me “[ð],” I would likely walk away from that interaction confused, and no
information would have been conveyed. (This isn’t to dismiss the value of
individual phonemes devoid of their contexts, but merely to point out that the
type of information conveyed by an isolated phoneme (or allophone) is not the
type of meaning that is exchanged in conversation. In general, throughout this
post, when I refer to a phoneme as holding no meaning, this is the distinction
I’m referring to.)
. . .
The phoneme [ð] on its
own does not convey much meaning. Add a schwa and you have [ðə] (the), which is
slightly more meaningful. Continue adding phonemes and eventually you might end
up with something like [ðə gəɹl ɹʌnz ənəˈbæʃɛdˌli], which conveys an entire
idea. Break that sentence down into its component parts, and it begins to lose
its meaning. Discrete morphemes such as [li] and [ən] still hold some form of meaning.
But continue breaking it down until you have, rather than morphemes, only
individual phonemes, and the meaning all but disappears. Neither [ɹ] on its own
nor the random fragmented sequence [ɹ], [w], [ʌ] conveys to the listener that
there is a girl, or that she is running, or that she is doing so unabashedly. .
.
We have all these theoretically discrete units of
language, of varying sizes and qualities. And it’s often tempting to look at
them as if there were an entirely linear, additive progression from one to the
next – phonemes combine to make morphemes, which in turn combine to make words,
which combine to make clauses, then sentences, then paragraphs, and so on. And
while in many senses this is true, it in no way tells the whole story;
personally I think that viewing language acquisition within this linear
construct is limiting and flawed in several ways, two of which I’ll discuss in
this post.
Firstly, language epitomizes the idea that “the whole
is greater than the sum of its parts.” Take the phonemes [ʒ] and [ə], for example. On their own, neither holds
any interpretable meaning; and yet when you combine the two you end up with the
French “je” [ʒə] (meaning “I”), which
conveys a tremendous amount of meaning. Those two phonemes, when combined in
that particular order, imply the entire concept of a self as distinct from the Other.
And it’s not that [ʒ] contains
half the concept of self and [ə] the other half; it’s that neither of those
phonemes holds any meaning until they are placed together, and, upon combination,
that meaning is suddenly created. When discussing language, it’s not as simple
as 1 + 2 equals 3. Rather, 1 + 2 equals 3 plus some other, less tangible
element which arises as smaller units are combined into larger ones.
The
second idea that isn’t quite captured by a systematic, linear view of the
process of constructing language relates to how we (humans) learn languages. As
an infant, the first type of language you are exposed to is probably comprised
of entire words and sentences, spoken between the adults around you; in other
words, before you are exposed to discrete individual phonemes, you are exposed
to the larger units of language that those phonemes combine to create. Yet the
first units of language that an infant actually produces are most likely individual phonemes and morphemes – an infant
produces these accidentally, while babbling, and later learns to combine them
in meaningful ways. And while these are the first units of language that an
infant will actually produce, they
are most likely not the first units
that he/she will understand. Most
children raised in an English-speaking environment learn to use the word “mama”
(or some derivation thereof) before they learn that the phoneme “un” usually implies negation; most children
raised in a Spanish-speaking environment will understand [mi’xita] (or [mi’hita])
(“mi hijita” [sometimes written as “mijita”] or roughly “my little daughter”) before understanding that [ita]
(or [ito]) is a diminutive suffix used to indicate smallness, cuteness, or
affection. . . .
It’s
also interesting to examine this a-linear acquisition of language in the
context of foreign language learning, and how this differs from the process by
which an individual learns his/her native
language. When you learn a second language as an adult (or as an older child),
you’ve already learnt to speak – that is, to produce organized combinations of
sound that, together, convey meaning. So you don’t usually begin by pronouncing
the individual phonemes and morphemes that exist within a language. Instead,
you often start with some common phrases. For example, if learning Spanish as a
second language, you might learn to say “Tengo
treinta años” ([teɪŋgo tʀeɪnta aɲos]) before you learn that “tengo” means “I have” and “años”
means years – and before you learn to pronounce [ɲ] as an isolated phoneme. After learning some
phrases you go back and learn individual words, and maybe as you learn grammar
you then break those words into morphemes in order to better understand their
meanings. And maybe at some point you work on pronouncing the individual
phonemes that exist in your second language and not in your native one. (All of
this is a very individualized process though; clearly this is just one of many
possible orders in which one might go about learning a second (or third or
fourth) language.)
So,
acquisition of a foreign language seems to follow just as winding and a-linear
a path as acquisition of one’s native language. Yet it isn’t the same winding path. It begins in a very
different place than does the acquisition of one’s native language. Rather than
first learning to produce phonemes and then learning to combine them in
meaningful ways, when learning a foreign language one begins by learning to understand
and pronounce (albeit badly) much larger units of meaning before then breaking
them down into their component parts. And I’m not necessarily saying that this
difference is a bad thing. Learning a language at birth and learning a language
later in life are fundamentally different – you begin each with different
knowledge and a different skill set, and so the process must necessarily be
different in order to accommodate that.
But
I think it’s interesting to look at the differences between the two acquisition
processes (and to look at each process in isolation as well) in conjunction
with the idea of language as comprised of a series of layers of ‘building
blocks’ (phonemes, morphemes, words, sentences, etc.) which combine to create
meaning.
. . .
It appears (at least to me) that learning a
language is not a linear process wherein one begins with the smallest distinct
units of sound (phonemes) and then combines them into bigger and bigger units
until eventually arriving at flowing speech. Rather, it’s a process wherein all
the different components of language are learned and relearned continually and concurrently,
starting in infancy and continuing until death.
Alden's post is very thought provoking because it makes me reflect on how I learned English as a young lad. Admittedly, my recollection of the events is rather fuzzy but my instincts tell me, as Alden suggests, I focused primarily on understanding sentences and words in the early stages of language building and seasoned my knowledge with a mastery of morphemes in the latter stages. Having a firm grasp of morphemes is where the real power lies in language utility because therein lies the power of getting your point across. As a kid when I said, "Iwa go-pa ahn swee" it took weeks before my mom understood that I wanted to "go to the park and play on the swings." Instead, according to my mom, I was constantly taken to the potty to pee. Disgruntled (and thoroughly dehydrated) I worked diligently to master the morphemes of the sentence, "I want to go to the park and swing" to get my point across. Consciously learning phonemes, however, is a different story. I certainly practiced phonemes diligently in the foreign languages I've studied but in English, I can't say I've given them much thought until this class. So my take is that as native speakers we put a great deal of intellectual energy into all of the components of language building EXCEPT phonemes. Those, I believe, most native speakers take for granted because they come intuitively. Only when studying a foreign language as an adult do phonemes get our intellectual attention because we do not have an intuition to guide us.
ReplyDeleteI would argue that there are not two but three methods of language acquisition, or, rather, that there are two distinct methods of foreign language acquisition. The first is “artificially,” through formal classes and instruction, and the second is “naturally,” being thrust into an immersive experience. I have experience with both: I learned Portuguese through immersion and French through formal classes and felt a distinct difference.
ReplyDeleteAn immersion experience places the focus on listening, understanding, and subconsciously training the ear to differentiate between the phonemes of the foreign language. Both vocabulary and structure are learned implicitly through context and the meaning of the words feel more intrinsic, in my experience. On the other hand, a formal class experience places focus on speaking, reading, and writing, while learning phonemes consciously rather than naturally. Vocabulary and structure are learned explicitly and based around the students’ mother tongue rather than context.
Between the two, I gained different levels of linguistic competence in different areas. I have a more extensive functional vocabulary in Portuguese, but I find it easier to communicate my thoughts in French due to structured speaking practice. Conversely, I have some difficulty understanding native speakers of French but can understand native speakers of Brazilian Portuguese with ease. Also, since I learned French after Portuguese, I was able to pick up French faster due to more experience with language learning.
So, to add on to your point, I find it interesting to look at the different forms of second- (or third-, etc.) language acquisition and how experiences with each language differ from and build upon each other, within the context of a non-linear process.
I agree that isolated phonemes are probably not how we learn to use language, but are obviously critical to our usage of language in order to communicate the ideas. The phonemes are the means by which we produce the ends of language, but I would argue that, to a certain extent, they must also be individually learned. Many foreign language classes teach the alphabet of that language, in an attempt to teach the learner how to make sounds in that language. Some languages are extremely particular when it comes to accuracy, and incorrect sounds can change the meaning of the word, or render speech unintelligible. Young children will mispronounce words, despite having the correct word and the correct meaning in mind. If the sounds are wrong, the word is wrong. This coordination between mental states and phoneme production is something that can only be fine-tuned by language use; thus, I would argue that the learning process is indeed linear. There must be some basic mastery of the sounds before the language can be comprehended in its spoken form. Although the sounds do not convey meaning, a speaker who has less proficiency with the sounds will inherently be less proficient in speech.
ReplyDeleteWhen you speak to a young child, even in long sentences, they often can understand everything you’re saying and nod, cry or let out some other response. But, as Alden touches on, they often cannot respond in anything but individual phonemes. But when I consider how toddlers speak, I am led to believe that this style of speech comes not from inability to string sounds together, but inability to make all sounds in the first place. How do we know where to place our tongue and lips to make the perfect m sound? When toddlers say mama, as opposed to mom or even mommy, perhaps it is because it’s simply easier to say given that it is comprised of two repeated syllables. You can’t say a sound you don’t know how to consciously make, so as children figure out how to express every sound, their speech becomes more fluent.
ReplyDeleteI was really intrigued by all the discussion of second language learning going on here. For me, learning Spanish jumps to mind. When I first started taking Spanish, every conjugation was extremely deliberate. It was only when I’d obtained some faint level of fluency that I stopped thinking of words as a root plus a conjugated -ar or -ir/-er ending.
ReplyDeleteWhen you reach this point in learning a language, the words become whole. You can say them on instinct, and they’re grammatically correct. They just sound right. This is the way all of us learned our native languages. I could draw up a conjugation chart for Spanish verbs much more easily than English verbs, but I wouldn’t need to pause for breath before using the correct form of the English word, because I’m far more fluent in English.
I'm right with Alden's idea of whole vs. parts up to the point of his dissection of secondary language acquisition. I think the main differences between "native" and "acquired" languages are perspective and environment rather than the order in which we learn them.
ReplyDeleteI moved to China in my early childhood, having spoken English here for a few years. Because I was so young, I forgot English and picked up Chinese as my primary language. When I was in an English class in China, the curriculum heavily focuses on grammar and writing. On the first day of class, the teacher drew comparisons between Chinese and English, pointing to the position of a subject, verb, object in both languages. For a lot of English learners, books were the primary means of interacting with the language. Having conversation with a native speaker can be a luxury, which means exposure to the spoken language can be very limited.
Language influences thought. A infant, having no prior frame of reference, can only piece together the world by being immersed in a language. An adult, however, is very inclined to relate a language to his prior knowledge. Additionally, access to different forms(spoken, written, lexicon) of a language is a resource that shapes the acquisition process of a secondary language.
This is a very interesting topic because of how we learn languages: in a formal classroom or through immersing ourselves in the language itself. I think the idea of learning phonemes usually takes place in the classroom as many before me have mentioned. The distinction between both learning environments is the need for the language. When you travel to Peru for example, your needs are about getting food, clothes, or to a destination. Thus, you are less quick to delve into the phonemes of a word since your need is more immediate. This changes in the classroom. I also wanted to touch on how phonemes themselves, as Alden mentioned, do not have meaning alone. They are merely sounds that as a society we have given meaning when piled with other sounds, body language, and context. I do think that we fine tune these sounds as we grow older because we now need more people to understand us. As a baby, you might get away with murmuring some version of "hungry," but as you age people might not understand which forces you to fine tune those sounds.
ReplyDeleteWhile reading this blog post and the subsequent comments, I found myself wondering if there are any languages in which all of the words are just single phonemes. In other words, are there languages in which a single phoneme like [ð] does contain meaning on its own. I have not heard of any such thing and cannot think of any such language, raising the question of why this is so. Why do languages consist of primarily multi-phoneme words? Does it have to do simply with the fact that we cannot make enough sounds to account for all needed words? In that case, are there languages that have multi-phoneme words, but in which all of the single-phonemes also have meaning on their own. I am wondering if someone who has taken other linguistics classes or is more familiar with this area of linguistics might be able to address this question.
ReplyDeleteI find it incredible that babies are able to simply listen to constant streams of sounds and eventually learn to parse them into distinct phonemes and morphemes. I also find it interesting that it is this step that seems to baffle cognitive scientists: the manner of parsing morphemes from speech and then assign meaning to these morphemes. Somehow, it is this very process that scientists are unable to code into machines.
ReplyDeleteI think the discussion is really hitting the nail on the head with regard to the “natural” process of acquiring the primary language rather than the “taught” process of foreign languages. This always reminds me of a personal passion of mine: music. Victor Wooten, one of the best electric bass players alive, is a proponent of the “music-as-a-language” theory. Wooten learned music the way we all learn our primary language: through experimentation and early exposure, as opposed to being taught. I find this parallel between language and music enlightening, and I believe that consideration of music acquisition can offer an interesting perspective on the study of language acquisition.
Alden claims that the non-linearity of language learning can be observed in early language acquisition. According to Alden, infants learn whole words and phrases before they learn the function of particular morphemes. Alden’s claim, however, rests on the assumption that “learning” is conscious and deliberate. If we expand the definition of “learning” to encompass unconscious and sub-conscious learning processes that occur outside of the learner’s awareness, we would find many studies suggesting that infants do in fact make use of phonemes early on in their life, even without having a firm grasp of the higher-level “meaning” that phonemes strung together can comprise. While a phoneme in isolation arguably contains no meaning, the statistical relationship between adjacent phonemes is meaningful in the sense that infants use it to learn how to segment human speech. The basis of so-called “statistical learning” in early language acquisition rests on the observation that more common phoneme pairs occur within words (i.e. in the middle of words), while less common phoneme pairs occur between words. In Saffran et al.’s (1996) seminal paper on statistical learning in early language acquisition, 8-month-old infants were able to segment words in a novel, artificially-created “non-sense” language they had never heard before by exploiting such statistical properties. (The paper can be found at this link and should be accessible if you have a Stanford account: http://goo.gl/nXcSGS. A summary of it can be found here: http://goo.gl/vxHQFW.)
ReplyDeleteI agree that on a conscious level we do not initiate our mastery of language through understanding of phonemes. However, at some sort of level -- perhaps beyond our initial awareness -- our knowledge of these plays a vital role in acquiring language. Thus, piggybacking off of Nathan's comment, a baby or person learning a new language must have mastery of individual phonemes before he or she can begin to produce full-fledged language. The example that comes to mind is baby talk, in which the baby is simply babbling different phonemes like 'ma', 'da', 'uu' and so on. Later on in development the baby is able to put these different phonemes together to produce a word and this later evolves into mastery of language. Therefore, I agree that phonemes may not seem to be the most important factor in our mastery of language. Regardless, I do believe acquisition of language is a linear process even thought it may not seem obvious at first.
ReplyDeleteI agree with the notion that language is not linear, but I disagree with the notion that language processing is nonlinear. What strikes me as fascinating, is how the brain can use these single, meaningless phonemes effectively for different languages. For example, "I, I, I" in English has almost the same pronunciation as "Ay Ay Ay!" in Spanish. What's interesting to me is that if I were listening to a recording and I wasn't told what language it was going to be in, my brain would be probabilistically trying to decrypt the sounds into firstly a language, and then into meaning. This introspective insight leads me to posit that language processing is parallelized, alluding to the concept of neural networks and neuronal activation. This, in turn, allows for the tackling of the language processing problem to be efficient.
ReplyDeleteMoreover, I do believe that language is non linear because of the very features Alden describes. A word has infinitely more meaning than a letter. Not only this, but with a finite set of letters / phonemes, we can produce an infinite amount of meaning. In itself, this is evidence that the semantic value of language is not linear.
As I read this post and then read the responses what I am seeing and continue to be uneasy with is this idea that language learning especially amongst children and those learning a new language is a singular process. That is to say that there seems to be this assumption that learning language is the only process occurring when a language is being introduced. And while that might possibly be true if you are learning a language in a university setting and only study inside a classroom that leaves out language immersion and the entirety of language learning by children. When learning a language as a child you are also trying to master walking and getting your needs met and communicating using body language or other cues. So therefore you are able to learn about language without this idea that it is only coming from phoneme breakdown or from understanding sentences. You are learning to match words to visual cues then breaking down those visual cue and language representations with the body language responses from parents and other children. There are many ways that language learning seems to be involved in not just learning what words represent what things and how those words make sense and sentences. Suffice to say learning a language is a complicated process, but acting as if it occurs in a vacuum or as a singular activity seems wrong and incomplete
ReplyDeleteI agree with Alden that viewing language acquisition as a linear process is flawed and should be viewed as a continuous process of learning on both smaller and larger scales. Although I agree that I first learned the meaning of words by listening to my parents speak before learning how to produce all the phonemes in the English language, I learned some English words in the same instructional manner as learning a second language. For example, the first comprehensible “word” that I spoke as an eight-month-old infant was “dada.” My parents and relatives “coached” me for months by repeating the word until it occurred to me that they wanted me to repeat it. However, I still had to figure out how to manipulate my articulators in order to reproduce the words that my parents spoke. Perhaps I didn’t even understand that “dada” referred to my dad, but I nonetheless learned an English word. I also don’t think that I consciously learned to produce the English phonemes through experimentation before learning to combine them to form words (as I wasn’t a talkative toddler), but rather by being “taught” by my parents to produce particular words at least during my infancy. It was later that the simultaneous processes of learning phonemes and putting them together to form comprehensible thoughts started after this initial phase of learning English as if it were a second language.
ReplyDeleteWith respect to your description of L2 acquisition involving learning larger sentences and phrases and only later learning individual morphemes, I have very strong feelings about this.
ReplyDeleteI personally believe that learning phrases rather than morphemes and grammar contributes to very ineffective language learning. When people learn phrases, they do not learn the morphemic building blocks that allow for people to formulate a wider variety of their own ideas and expressions in the L2. In your example, learning “Tengo treinta anos” allows one to communicate a very narrow idea, their age. Learning that “tengo” means “I have”, and learning the conjugations of “tener” more generally allows one to communicate about ownership in a huge variety of contexts.
I think that, because Latin students learn morphemes and grammar from day one of class rather than fixed expressions, they are able to take on much denser texts at the 4th year AP Level (Vergil, Caesar) than are students of French or Spanish, though I am admittedly biased as a former AP Latin student.
In principle, then, I believe L2 learners should learn a bit more like L1 learners and focus on putting together morphemes rather than memorizing phrases.
Building off of Paul Gregg, I also think music is interesting to compare to the other spoken languages mentioned. Like spoken languages, music is communicated through sound.This leaves a lot of speculation to the listener for understanding music. Sure, some music may be straightforward with lyrics that can be understood, but otherwise music is most definitely a language where the meanings of the bits and pieces haven't really been defined. On the topic of definition, I don't know if music could ever be defined in such a clear-cut way.
ReplyDeleteAlthough thoughts can be communicated through music, they are much harder to associated with the music, and harder yet to keep consistent. I feel that viewing it as a language keeps it open-ended and interesting.
Even though I mentioned earlier that we don't have a clear way to understand the thoughts associated with music, we are fortunate enough to have notation for it, a written form of the language. To some extent this notation may be able to give some information as to how to perform the music, but still no insight to understanding it.
Music may be a language that can be learned in a similar way to other languages, starting with pieces, and learning how to weave them together into a single phrase or idea. In this way, even though very different from most languages involving sound, music doesn't seem too far off from a system we have already learned.
I really enjoyed this post, especially because I've had different experiences learning different languages throughout my life. I learned Korean as my native language, English when I was 9 years old, and Japanese when I was around 16. After reading this post, I realized how different each of those learning experiences was for me.
ReplyDeleteOf course, when I first learned English, the language did not come naturally to me. For a couple of years, whenever I had to say something in English, I thought of it in Korean first, and word by word, translated it into English. Then, I applied the different grammar rules, shifted the words around, and (finally) spoke out loud.
The difference between learning Korean (native language) and English (second language) was that I didn’t have a reference to look to when I was learning my native language. I didn’t have a process where I thought of what to say in another language, and then translate to Korean. It was something I had to do from scratch. Therefore, as Alden mentioned, I do agree that learning a language is not a linear process.
Another example is when I learned Japanese. Japanese grammar is pretty much identical to Korean grammar. However, I know more vocabulary in English than I do in Korean. Thus, I learned Japanese grammar by comparing it to Korean grammar, and Japanese vocabulary by comparing it to English vocabulary. If you tell me to translate something from Korean into Japanese, I’ll have to translate the Korean words to English, and then to Japanese; and when you ask me to translate from English to Japanese, I’ll have to think of the grammar in Korean first. Weird! My third language was acquired by comparing it to both Korean and English.
From my experiences, I wholeheartedly agree with Alden’s last sentence: “[Learning a language] is a process wherein all the different components of a language are learned and relearned continually and concurrently.” Now, when speaking to my parents, I mix English and Korean together unconsciously (a.k.a. Konglish)! My dad understands Japanese as well, so sometimes I mix all three of them together... it’s really funny how learning each of the languages was so vastly different but also interlinked.
Sophia’s comment was very interesting to me, and I’ve noticed similar trends in my own language learning. My first language is English, my second is Spanish, and my third French. When I was first learning Spanish, the process was similar to what Sophia describes – I would think of things in English and then translate them into Spanish before speaking. As I continued learning Spanish I began to think directly in Spanish, and no longer used English as an intermediary. Now, when someone speaks to me in Spanish, I automatically respond in Spanish; when I hear something in Spanish, I understand it without thinking of what it would be in English. The strange thing to me is that, after having lived in a Spanish-speaking country for a year, that process seems to have gone even further, or almost reversed itself. Now, when speaking to my English-speaking friends, there are many instances in which I will think of a sentence in Spanish and then have to mentally translate it into English before speaking it; or often when I am writing, certain words or concepts will occur to me in Spanish and I have to take extra time to think of how to express them in English.
ReplyDeleteAlso, I’ve found that, in learning French, it’s much easier for me to skip the intermediary translation step than it was when I first learned Spanish. In other words, even when I was only first learning French, I would hear (or read) something in French, understand it, and respond – and I wouldn’t mentally translate into English to do so, as I had done when first learning Spanish.
Sophia mentioned that the difference between learning her first and second languages was that she “didn’t have a reference to look to when learning [her] native language.” Related to this idea of referring back to one language when learning another, I’ve noticed that, on the occasions when I do still mentally translate before forming French sentences, I find that I translate between French and Spanish, rather than French and English; and the errors I make when speaking in French are almost always based off of Spanish grammatical structures rather than English ones. It’s as if Spanish (rather than English) is the reference point from which I’ve learned French. I don’t know if this could be because French and Spanish are so closely related (i.e., both are Romance languages), but it’s just something I’ve noticed . . .