The fields of linguistics


Please introduce yourself on our Facebook page if you haven’t already done so.


I would like you to keep track of the time it takes you to work through this chapter, so please look at a timepiece now.



I probably should add some sort of advance notice of what this chapter is about.

The fields of linguistics

I would like to start out by asking you a question.


What are the parts of the sentence below?

Flying planes can be dangerous.

Let’s start with the smallest.

Speech sounds

You would think that the smallest elements of the sentence are its letters. There are twenty-six of them, plus four spaces and a period. You might also think that the letters tell you how the sentence is pronounced, but is that really true? How is ‘n’ pronounced? How is ‘g’ pronounced? How is ‘ng’ in ‘Flying’ pronounced, like ‘n’+’g’? And while on the subject of ‘g’, is it pronounced the same in ‘dangerous’ and ‘Flying’? Turning to ‘e’, how is it pronounced in ‘planes’, ‘be’ and ‘dangerous’? While ‘a’ pronounced the same in ‘planes’ and ‘dangerous’, what about in ‘can’?

I hope that you are starting to grasp that individual letters have at best a tenuous relationship to the pronunciation of the words made from them in English. Or to put it more succinctly, English spelling is a mess. Consider yourself lucky that you learned it as a child, when you were really good at learning stuff.

It is not just English, though. No written language captures pronunciation perfectly. This makes it impossible for linguists to use normal writing to show how a word is pronounced, so they have had to invent a new alphabet, called the International Phonetic Alphabet or IPA for short. In the IPA, each letter represents a single speech sound, and each speech sound is represented by a single letter. Since there are not enough letters in the Roman alphabet to represent all the possible speech sounds, letters have been adopted from the Greek and runic alphabets, if not made up from scratch. As an example, the IPA representation of our sample sentence is as so:

'Flying planes can be dangerous.'
[flaɪiŋ pleɪnz kæn bi deɪnʤəɹəs]

I have enclosed both sentences in different delimiters to call your attention to the fact that they are different descriptions of the same object. The act of converting speech to written characters is called transcription. Single quotes ‘’ signal that the transcription is to English spelling. To be precise, the characters of English spelling are referred to as graphemes; a transcription with single quotes is called a graphemic transcription. Square brackets [] signal that the transcription is to IPA. The IPA characters are referred to as phones; a transcription with square brackets is called a phonetic transcription.


What is the phonetic transcription of ‘flies’, ‘ban’, ‘bank’ and ‘judge’?

What is the graphemic transcription of [næt], [ʤiɹ], [keɪn], [bəs] and [ɹiŋ]?

From now on, I assume that the smallest units of speech are IPA characters.

The physics of speech: phonetics

I have used the words ‘pronounce’ and ‘pronunciation’ several times now; what do these words mean to you?

What they mean is the process by which the mouth or vocal tract produce speech sounds. As an example, put one hand on your throat and pronounce ‘bus’. Now change to ‘buzz’. Change back to ‘bus’. You should notice three phases. Both words start with both your lips coming together briefly to stop air from flowing out of your mouth. This is the essence of the consonant [b]. Then your mouth opens up and something you can’t really feel very well happens to produce the odd vowel [ə]. Finally, the words end with your tongue coming near the back of your teeth and constricting the airflow to produce the consonants [s] and [z]. What is the difference between them? That’s why I asked you to put your hand on your throat: it makes a buzzing or vibration for [z] but not for [s], produced by your vocal cords.

This precisely timed ballet of lips, tongue, and vocal cords is termed articulation. Its study is known as articulatory phonetics. The point of the articulatory ballet is to induce a vibration in the air – studied as acoustic phonetics – a vibration that hopefully makes it to your ears, where it is converted into something that the brain can understand. This is commonly known as hearing, whose technical term in the health sciences is usually audition. For our specific concern with the audition of speech, the field of linguistics is called auditory phonetics.

We will talk about acoustic speech signals and their transduction into auditory objects in the next day of class. This will necessitate learning something about articulation, too, though the meat of articulatory phonetics is deferred until we reach motor cortex towards the end of the semester.

The production and perception of speech may be so strongly interdependent in neurophysiology that it comprises what is called a sensorimotor system, mediated by an acoustic signal. That is, the areas of motor cortex that control the vocal tract tell the vocal tract how to move to produce a phone, which is picked up by the sensory organ for audition – the ear and connected areas – and relayed to the corresponding sensory cortex.

The cognition of speech: phonology

For animals with a limited vocal repertoire, this about all there is. For humans, who have the largest vocal repertoire in the world, a lot of other stuff happens between primary sensory cortex and primary motor cortex. For lack of a better term, I prefer to call this other stuff cognition. To many people, cognition denotes thinking. I stretch it to include what is variously known as feeling, emotion or affect. I even stretch it to include the possibility of sensory and/or motor cognition. Speech provides a particularly clear domain for both, or sensorimotor cognition. Let us walk through two examples.

Articulate the words ‘dodge’ and ‘Don’. Is there any difference between them besides that of the final consonant, [ʤ] vs. [n]? For most English speakers, there is. The vowel [a] of ‘dodge’ is slightly different from the [a] of ‘Don’. It sounds a bit like [n], that is, it is nasalized. English vowels are normally articulated orally – in the mouth – but sometimes some air can flow out through the nose and give the vowel a nasal resonance. In the IPA, nasality is represented by putting a tilde [~] above the vowel, as in [ã], so that ‘dodge’ and ‘Don’ can be transcribed phonetically as [daʤ] and [dãn].


Decide which of the following words have nasalized vowels:

‘mop’, ‘mod’, ‘bone’, ‘bass’, ‘bond’, ‘bad’, ‘nab’

Can you think of a rule or generalization that predicts whether a vowel is nasalized or not?

The answer is important, so I will go over it here.

One way to grasp what is going on is to organize the data into a table according to where a nasalized vowel could be found with respect to the surrounding consonants. In our small corpus of three-letter words, there are three main possibilities, where “N” stands for any nasal consonant, “O” stands for any oral consonant, and “V” stands for any vowel:

Table 2 Three contexts for vowel nasalization in English




‘mop’ [map]

‘dodge’ [daʤ]

‘Don’ [dãn]

‘nod’ [nad]

‘bass’ [bæs]

‘bone’ [bõn]

‘nab’ [næb]

‘bad’ [bæd]

‘bond’ [bãnd]

Organized this way, into just three contexts, the one that predicts nasalization should jump out at you. It is that a vowel is nasalized in English if it is followed by a nasal consonant, which in English are [m, n, ŋ]. It will be handy to have a name for this generalization, so I hereby dub it the English Nasalization Rule.

Other languages take nasalization much more seriously. In the table of four French words below …

Table 3 Oral vs. nasal vowels in French

















How does nasalization of a vowel affect a word in French?

Can you think of a rule or generalization that predicts whether a vowel is nasalized or not?

Well, the English rule doesn’t work, because the nasalized vowels are not followed by a nasal consonant. The consonant before the vowel does not appear to have any influence, either, so our conclusion must be that French vowels are not nasalized by anything in their context – they just are. In other words, some vowels in French are purely oral, others are nasal, and the only way that you can know is by learning it word by word.

Another way of explaining the difference between English and French vowel nasalization is that, in English, nasalization does not have the ability to change the meaning of a word. Any of the English words with a nasalized vowel in the third column of Three contexts for vowel nasalization in English can be articulated without the nasalization without changing its meaning. In contrast, if you articulate French [sã] without the nasalization as [sa], well, you have said “his” or “her” instead of “without”.

The crucial thing is that nasalization is physically identical in both languages. Exactly the same acoustic waveform is articulated and perceived by French and English speakers. What is different is how each language organizes this waveform, an instance of sensorimotor cognition. In English, nasalization can be predicted by its context; in French, it cannot and so has to be memorized as part of a word. This difference is so profound that linguists have developed yet another nomenclature and notation to state it precisely. A phonetic transcription can be reduced to just those phones which are not predictable from their context, called phonemes and represented between forward slashes // as a phonemic transcription. The following table illustrates this notation:

Table 4 Phonetic vs. phonemic nasalization
















The phonetic transcription describes what you hear and articulate; the phonemic one describes what you think. The former makes up the branch of linguistics known as phonetics, with the three sub-branches already mentioned, articulatory, acoustic and auditory. The latter makes up phonology. We will come back to this distinction once we get beyond auditory cortex.

There is a bit more terminology. The units of phonetics are called allophones or sometimes just phones; the units of phonology are called phonemes. Since the choice between two allophones of the same phoneme can be predicted by context, the two allophones are said to be in complementary distribution. The occurrence of two phonemes is not predictable but rather must be memorized as part of the word, so they are said to be in contrastive distribution.


The following table summarizes the differences covered in this section:

Table 5 Phonetic vs. phonemic organization
















Returning to the task which started this chapter off, that of dividing a sentence into its parts, we can now clarify what the finest division looks like:

‘Flying planes can be dangerous.’ [flaɪiŋ pleɪ~nz kæ~n bi deɪ~nʤəɹəs] /flaɪiŋ pleɪnz kæn bi deɪnʤəɹəs/

Additional practice


The parts of words

Recall that the question that we are trying to answer is what the parts of “Flying planes can be dangerous” are. We just discussed the smallest parts, those that correspond to individual sounds. The next level is to combine sounds into word parts. Which words in the sentence seem to have more than one part?

Let me organize them into a table:

Table 6 Complex morphemes in the sample sentence













/flaɪiŋ/ appears to consist of the verb /flaɪ/ plus the suffix /iŋ/ whose exact category I don’t want to go into yet. /pleɪnz/ appears to consist of the noun /pleɪn/ plus the suffix /z/ which makes it plural. /deɪnʤəɹəs/ appears to consist of the noun /deɪnʤəɹ/ plus the suffix /əs/ which makes it an adjective. /kæn/ and /bi/ don’t appear to be decomposable into anything smaller.

In linguistics, these parts are called morphemes; the field of linguistics which studies them is called morphology.

As the examples demonstrate, much of the morphology of a word depends on its part of speech. It you are rusty on the parts of speech, The Eight Parts of Speech will jog your memory.


Divide the following words into morphemes. Can you tell what each morpheme does?

‘morphology’, ‘blackberry’, ‘cranberry’, ‘transport’, ‘blacken’, ‘disassemble’, ‘run’

What is the longest word in English? Post your longest word to our Facebook page. Divide it into morphemes, on your own, not necessarily on Facebook.

Putting words together

So we have finally reached the level of the whole word. You may think that we are finished, or at least that the only thing left is the sentence, but not so fast. What does our sample sentence “Flying planes can be dangerous” mean? To give you a hint, it can be continued in two very different ways:

  1. Flying planes can be dangerous. That’s why I would never fly one.

  2. Flying planes can be dangerous. One crashed just the other day.

It can also be restated, or paraphrased, in two different ways:

  1. To fly a plane can be dangerous. That’s why I would never fly one.

  2. Planes up in the air can be dangerous. One crashed just the other day.

I hope that I have made it clear that the same sequence of words can have different meanings or readings, which is to say, that it is ambiguous. The ambiguity here hinges on the form of ‘flying’. In the first reading, the ‘ing’ suffix turns the verb ‘fly’ into a noun, and ‘planes’ is the direct object of the verb, telling us what is flown. In the second reading, the ‘ing’ suffix turns the verb ‘fly’ into an adjective, which tells us what kind of planes are being talked about.

Linguists have developed a notation to represent this difference, which involves surrounding words in square brackets and subscripting a label to the first one, like so:

  1. \([_{NP} [_{ADJ} \text{ flying}] [_{N} \text{ planes}]]\text{ can be dangerous.}\) That’s why I would never fly one.

  2. \([_{NP} [_V \text{ flying}] [_{N} \text{ planes}]]\text{ can be dangerous.}\) One crashed just the other day.

In both cases, ‘flying’ and ‘planes’ are combined to form a larger unit, a noun phrase or NP, that is the subject of the sentence. But how do I know that?

You could reply that is is because grammarians say so, but linguists contend that there should be evidence for such a claim. And there is. Consider how ‘flying planes’ acts like a unit in the examples below. How do the (b) sentences sound in comparison to the (a) ones?

    1. What can be dangerous? Flying planes.

    2. What planes can be dangerous? Flying.

    3. but: What planes can be dangerous? Flying ones.

    1. What can be dangerous is flying planes.

    2. What planes can be dangerous is flying.

    1. It is flying planes that can be dangerous.

    2. It is flying that planes can be dangerous.

    1. Flying planes which are overloaded can be dangerous.

    2. Flying which planes are overloaded can be dangerous.

    1. Flying planes can be dangerous. They are everywhere.

    2. Flying can be dangerous. They planes are everywhere.

In the (a) examples, ‘flying’ and ‘planes’ are kept together, as a unit. In the (b) examples, ‘flying’ and ‘planes’ are separated. Since the (b)s sound considerably worse than the (a)s, it appears that the mind does treat the two words as a unit in this sentence.

While the effect of ‘ing’ on a verb is properly an aspect of morphology, the position and combination of words is complex enough that it deserves its own field of linguistics, that of syntax.


Check out the post Crash blossoms at the Language Log, and especially the comments. Pick your favorite crash blossom – or use one of your own – and paraphrase it on our Facebook page so as to explain its meaning(s).


To recap briefly, we have scaled a hierarchy of linguistic units that goes from phones to morphemes to words to phrases to sentences. You may think that this was done in a strange way, because almost no mention was made of the meaning of the units at each level. This is because there is some controversy in linguistics about how meaning interacts with the construction of a sentence. One school, called Generative Grammar, holds that meaning is ‘read off’ a fully constructed sentence. For the other school, called Cognitive Linguistics, construction of a sentence is really construction of a meaning, and there is no independent existence of morphology or syntax.

About the only thing that everyone can agree on is that the study of meaning is called semantics, and that it is not clear that there is any elemental unit of semantics, because a morpheme is usually considered the smallest linguistic unit with its own meaning.

Meanings is such a slippery concept that it can be studied neurolinguistically in only the crudest of ways. In particular, the sorts of meanings that are usually examined are those that hold between isolated words, as in the list below.


Try to identify the semantic relationships that hold between the words of each line:

  1. violin, fiddle

  2. hot, cold

  3. finger, hand

  4. wood, chair

  5. virus, epidemic

  6. car, plane, train

  7. I stowed my gear in the trunk. vs. I stowed my gear in the asparagus.

One of things that makes semantics so hard to study is that it often depends on a context. The classical example is of a mother and a child in a room together. The child says, “I am hot”, and the mother brings her a glass of water, turns on a fan or the air conditioning, or opens a window. Does “I am hot” mean “bring me a glass of water”? No, it doesn’t, but in a cooperative relationship, the expression of a negative state of being by one person can be understood by the other as a request to ameliorate the negative state. This has nothing to do with the literal meaning of “I am hot” but rather with how such a statement can be used in a certain context. The study of “how to do stuff with words” is called pragmatics.

Summary of the fields of linguistics

There are several other fields of linguistics, but I want to leave time to introduce neuroscience, so I will just summarize the ones that are most relevant to this course in the following table, with some observations about the others:

Table 7 Fields of linguistics




the study of writing systems; a little with visual language

auditory phonetics


articulatory phonetics





the study of stress and intonation; yes








a little


the study of words; implicit in some of the others

language development

the study of how children learn language; maybe

bi- or multilingualism

the study of people who speak more than one language; a little


the study of how language varies by social group; no


the study of the language of specific (usually geographically defined) social groups; no

historical linguistics

the study of how languages change; no, neuroscience can’t study dead people, but …

evolution of language

the study of how humans acquired language; maybe – what fun! almost pure speculation

anthropological linguistics

the study of how language varies by social group, usually pre-industrial or non-Western; no



Please report how long it took you to get here on our Facebook page by adding a comment to the relevant post.


  • Poeppel, D., Emmorey, K., Hickok, G., & Pylkkänen, L. (2012). Towards a new neurobiology of language. The Journal of Neuroscience, 32(41), 14125-14131.

Powerpoint and podcast

  • Aug 28 (W), day 2: BrLg-02-FieldsLing.pptx in Canvas → Files → ppt

  • Aug 28 (W), day 2: BrLg-02-FieldsLing.MP3 in Canvas → Files → mp3

The next topic

Induction (Philosophy), Models, Frequency; which I am still working on.

Last edited: Aug 22, 2023