Phone Home

I get a ton of emails. I mean, a ton. I have several email accounts, and it’s a part-time job to keep up with them all. Of course, nowadays, I also access email on my phone. I know I am not alone in this. Needless to say, a lot of the emails I get are language questions. Here’s one I got this morning, and I decided to turn it into a LEX Q&A, so more people can benefit from the dialogue than just us two. (The email has been edited for formatting and asides).

[W]hat is the final phoneme in the word cat when it is at the end of a sentence?  “I saw a little cat.”  It’s not the same as at the beginning of tip, but is it just an allophone of /t/?   I was reading about the “flap” and it doesn’t seem like it would be a flap, because my tongue stops on the roof of the mouth rather than tapping there. But I’m not sure how the flap works either. I feel as though when I say little I go straight from /ɪ/ to /l/. But there’s a difference between the way I say little and Lil. If I try to say Lil as a two syllable word with just the /l/ in the second syllable that’s still not the same as little so something is happening with my tongue, but I can’t figure it out. It almost feels like I’m squishing air out of the sides of my mouth in between the /i/ and /l/ and pushing my tongue more forcefully up with the final /l/ in little.

Aaaaaaand, my response: What a great question! And an important one, too. One of the biggest problems with the decades-old emphasis on “phonemic awareness” is that most teachers don’t really understand what a phoneme is. They think it’s a “minimal unit of sound” or some such; it’s not. It is minimal, and it is a unit, and it does have to do with language as it is pronounced, but it’s not actually a sound. Moreover — and this is critical — it’s distinctive. What this means is that, while it carries no meaning itself (the /b/ in /bɪt/ doesn’t mean anything), it is distinctive for meaning — it differentiates meaning — from other phonemes (the /b/ in /bɪt/ and the /p/ in /pɪt/ distinguish the meanings of those two words. That all happens in your head.

Elsewhere, however, there are different physical realizations of pronounced words and utterances. Those physical realizations have structures that can be studied, like all physical things. The phoneme /t/ is conceptual, a psychological category, container, or class — choose your metaphor — with several different possible members. Those members — all the members of the phoneme /t/ — are its allophones. Some physical realizations of /t/are aspirated. That is, they have a little release of air when the tongue is released from the roof of the mouth. That’s like in the word top. Phonemically, we would represent this as /tɑp/, but phonetically, it’s [tʰɑp]. If we put a /s/ in front of the word, however, the aspiration isn’t there: [stɑp]. You can see and feel the difference if you pronounce those two words aloud while holding a kleenex in front of your face. But phones aren’t necessarily distinctive for meaning: if you were in my car and yelled [stʰɑp], I would totally slam on the brakes. The [] and the [t] are allophones of the same phoneme, /t/. Other allophones of /t/ in English include [t ̚ ], [ʔ], and [ɾ], also known as the “flap.”

So, to answer your question directly, the phoneme at the end of cat is the same as the phoneme at the beginning of tip, but they are different phones. They are phonologically the same, but phonetically different. Yes, that makes them allophones of the same phoneme, different members of the same class.
Another allophone is the flap [ɾ] in your pronunciation of little. A Brit would be likely to say [lɪtʰəl], while an American more likely to say [lɪɾḷ]. The difference between Lil and little is that flap — your tongue briefly taps the alveolar ridge, before releasing the [l] laterally. There’s a co-articulation from the [ɾ] to the [l]: both of them have an alveolar place of articulation. You don’t have to move your tongue to get from one to the other. They are also both voiced. The difference between them is in their manner of articulation: [ɾ] is a flap, and [l] is a lateral approximant. That lateral refers to the release of the air out the sides of your tongue, just as you articulated in your question. The “more forceful” push of your tongue to the alveolar ridge in little? That’s the flap.

Phones and phonemes are not for sissies, but a clear understanding of the difference is absolutely critical for scholars and teachers of the written word. Writing systems’ representations of pronunciation may target syllables, or it may target phonemes, or both. But spelling never, ever targets phones; there’s no such thing as a non-phonetic word, or rather, all written words are non-phonetic. When a child writes <chree> instead of <tree>, she’s not mishearing the word; she’s ascribing the physical phone she is saying or hearing to the wrong phoneme in her head. *That’s* phonemic awareness, but teachers may be at a loss to remedy it unless they have clarity about what’s going on phonetically in that word.

No pithy ending in this post, no clever turn of phrase. No LEXlover’s delight. What do you want from me? It was an email. If you’re still reading this far, good for you, and you’re welcome.

15 Comments

  1. Gail Venable says:

    Well, it had a pithy title!

  2. Maybe this relates to the fact that consonants have no voice unless accompanied by a vowel – try saying t without following it with a little grunt (a schwa) sound. ‘Consonant’ actually means ‘with vowel sound’, and therefore the pronunciation of consonants must surely depend a lot on accompanying vowel sounds. Meanwhile, I will re-read, soon, but this is my ‘quick take’ on your A to Q.

    • Thank you for your ‘quick take’, Paquita. I’d like to clarify a few things with a little more leisure:

      You said, “Maybe this relates to the fact that consonants have no voice unless accompanied by a vowel.”

      I’m not sure what you mean by “this” as your subject, but if you mean the topic of my post, I don’t see the connection. Consonants are not the only phonemes that can have several allophones — vowel phonemes have allophones too. The fact that phonemes have allophones — the point of this post — has nothing to do with their nature as either vowels or consonants.

      It’s also not accurate to say that “consonants have no voice unless accompanied by a vowel.” Consonants can be voiced or unvoiced, regardless of any vowel accompaniment. For example, /b/ is voiced, /p/ is unvoiced; /ʒ/ is voiced, /ʃ/ is unvoiced. Voicing has to do with the resonance of the vocal folds, not with vocalic accompaniment.

      You go on to say “try saying t [sic] without following it with a little grunt (a schwa) sound.” Well, I say [t] all the time in words without following it with any little grunts or schwa sounds. What you are trying to articulate here is that [t] is a plosive, or a stop: its manner of articulation requires a release. If pronounced in isolation (which is pretty artificial), yes, it is often followed by an aspiration or a brief vocalic release. But that’s not true of all consonants, at all. Here are some consonants that I can pronounce with no “accompanying” vowel: /m/, /n/, /ŋ/, /s/, /z/, /v/, /f/, /ʒ/, /ʃ/, and many more! That’s because they are continuants, not stops. Some of them are voiced, and some are unvoiced, but none of them require a vowel in order to be pronounced.

      Next up: your claim that “‘Consonant’ actually means ‘with vowel sound’,” not exactly. The orthographic denotation of consonant is ‘sounded together’ because folks talking about language in the Middle Ages were under the same misapprehension that consonants could only be pronounced with a vowel; in that sense, it’s a little bit of a misnomer.

      The pronunciation of consonants, like the physical realization of any phoneme, is affected by coarticulation, but not only with vowels. I gave you a few examples of consonant coarticulation in the post: the /t/ in track is coarticulated with the /ɹ/; the coarticulation of the /s/ and /t/ in stop results in no aspiration. So, when you say, “the pronunciation of consonants must surely depend a lot on accompanying vowel sounds,” that’s not untrue, but it’s not exclusive to proximal vowels.

      Ultimately, Paquita, I’m not sure at all that I see the point of your offering here. I’m not trying to be unpleasant, but I am protective of my blogspace, and I’m not interested in compromising rigor and accuracy in the interest of magnanimity. Perhaps after you re-read soon, you will be better able to explain your perspectives on the language and how they are germane to what I’ve explained in my post.

  3. Pithy no, but still exact and brilliantly said!

  4. Sue Hegland says:

    Wonderful, Gina. Thank you.

    This was very interesting: “Phonemically, we would represent this as /tɑp/, but phonetically, it’s [tʰɑp].” I’ve been looking up words in the OED or other resources that show pronunciation in IPA, and thinking that I’m seeing the representation of phones. What I’m gathering from your comments is that I’m actually seeing a phonetic representation of the most common allophone for that phoneme. Is that a fair statement?

    Do any references list a word like “top” as [tʰɑp]?” I realize that many speakers of the language might pronounce that word differently, but I’m just wondering if there’s anything out there that would be helpful for checking my own understanding of allophones of phonemes.

    • What you see in most dictionaries is mostly, but not entirely, phonemic. The thing is, the IPA can be used to represent both phones and phonemes; whether a phoneme is represented with a given feature (like the aspiration [ʰ] in [tʰɑp]) depends on whether that feature is distinctive in that language. So, for example, /p/ is a phoneme in English, /pʰ/ is not. However, in Thai, both /p/ and /pʰ/ are distinctive phonemes, as are /t/ and /tʰ/. That’s why Thai is spelled with an ‘h’.

      So, dictionaries are mostly phonological, but they miss a lot of finer points. My Mactionary, for example, lists /ə/ in stressed syllables, as in couple: |ˈkəpəl| — it should actually be /’kʌpəl/. A schwa is unstressed by definition in English. They also use a ‘d’ in words like little: |ˈlɪdl|, but there is no /d/ phone or phoneme in that word. There’s a /t/ phoneme that is realized as a [ɾ] allophone in American English, but as a [tʰ] in RP English. They hedge their bets whenever there are glides and liquids, as in actual: |ˈæk(t)ʃ(əw)əl| — this is neither phonological nor phonetic; it’s largely phonological with some phonetic possibilities thrown in for good measure.

      You might notice that the Mactionary uses | | brackets, not / / nor [ ]. They’s a cue that they are not doing linguistics. And frankly, in a day and age when learners can press a button on their cell phone and *hear* the pronunciation of a word, these arcane lexicographical symbol mish-mashes will continue to obsolesce.

      I know of no references that list phonetic pronunciations of words. That’s not the point of phonetic transcription, from a linguistic perspective. There are far too many variables. I would mistrust any resources that purported to offer what you are seeking.

      In the LEX deck, however, I do list or address allophones variation on several of the cards. I list ɾ on the ‘t’ card and ɹ on the ‘r’ card, for example. But I spell that out (ha! no pun intended) on the green instructions cards.

  5. Lisa Barnett says:

    I’m offering a simple wondering since my mind is whirling with some confusion (due to my lack of knowledge) and new thoughts after reading this blogpost, I am wondering what keyboard or font offers you the multitude of options for all the phonemes you type — I am actually now very hesitant to say ‘phoneme’ because I thought I understood what it was but now realize my utterly over-simplified understanding of a phoneme–and unfortunately I’m a few steps ahead of most in my pack on the topic of phonemic awareness….Hmmm… At any rate, there are a few symbols used in your post that I’d like to be able to look up and I’d like to be able to use some basic ones on the occasions in which I type some very simple pronunciation keys for my students’ parents or colleagues. Thank you for such an in-depth post, I’m learning a lot from your Orthographers FB posts, radio cast with DTI & website. I hope to take one of your classes on SWI soon.

    • Hi Lisa,

      I use exclusively the International Phonetic Alphabet (IPA) for my discussions of phonology, phonemes, and phonetics. I applaud you for pegging your own understanding of phonemes as ‘over-simplified’ although you outpace your pack; it’s not your fault, and there’s a great deal of facile dreck about phonemes out there, I’m afraid, starting wit the assumption that they can somehow be discussed / located outside of morphemic structure. They can’t.

      There are many, many resources out there for IPA. I don’t know why we don’t teach it in schools; it’s a very useful system to understand. I’d imagine you can google it and locate it online. You may be able to search for it in my blog to see what other posts are around. Perhaps I should offer a 1-day introductory LEXinar on the IPA. Would that be helpful?

  6. lynnheasley says:

    Thanks for all you do better our understanding of how our language works. I feel this work truly is the missing link to help our struggling readers and writers especially!

  7. leslie923 says:

    Thanks for a very clear distinction between a phoneme and a phone, Gina. I couldn’t have said it better. As an SLP grad student, we had to transcribe disordered speech, so that was all phonetic transcription. All in the ear, not the mind. As I was reading through the question posed, particularly about the pronunciation ‘little’ I was wondering if the asker of the question is pronouncing this word with a glottal stop. S/he writes:

    “I was reading about the “flap” and it doesn’t seem like it would be a flap, because my tongue stops on the roof of the mouth rather than tapping there. But I’m not sure how the flap works either. I feel as though when I say little I go straight from /ɪ/ to /l/. But there’s a difference between the way I say little and Lil.”

    If this person is using a glottal stop, the tongue tip might move to the alveolar ridge to anticipate the word-final /l/. I have a friend from New England who uses the glottal stop rather than a flap in many of her pronunciations, the word ‘little’ being one. I wonder if that’s what’s happening with this person.

    Also, isn’t there a diacritic for the phonetic transcription of an unaspirated stop consonant. Unfortunately I don’t have access to it on my keyboard, but it looks a bit like the numeral 7 without the angled decender.

    Thanks for the post!

    • Leslie, a flap needn’t be unreleased for it to still be a flap. But release, aspiration, glottal stops — these are phonetic concerns in English, not phonemic ones. These would only be represented in a phonetic (narrow) transcription, not in a phonemic (broad) transcription, and certainly not in the writing system. The phonetic transcription for the unreleased flap would be [ɾ ̚] but the phonemic transcription for the flap would be either /t/ or /d/, depending on the word.

      Also, you write about the apico-alveolar articulation of /l/, but we can’t ignore that the articulation of the glottal stop also involves the tongue: the root of the tongue, in the back of the mouth. We have to pay attention to the whole of the tongue, articulatorily speaking, not just the apex.

  8. MiriamSLP says:

    Thank you for such a precise and helpful explanation! Often there’s a lack of clarity about what phonemic awareness really is and understanding what a phoneme actually is a great start! Sometimes the terms “phonological awareness” and “phonemic awareness” are even used interchangeably, obscuring the importance of the phoneme level for the task of learning to read.

    I’m curious about something: I’ve just finished reading “Early Reading Instruction” by Diane McGuinness. Based on her study of writing systems, she contends that they are each based on one and only one of four phonological units: syllable (Chinese, Sumerian); CV diphone (Japanese hiragana and katakana, Han’gul); consonant cueing system (like Hebrew, which with the addition of vowel diacritics basically becomes an alphabet); and phonemes (alphabets).

    You write: “Writing systems’ representations of pronunciation may target syllables, or it may target phonemes, or both”. Do you have an example of a writing system that targets both?

    Thanks!

    • A good exemplar for a writing system that targets both phonemes and syllables is Han’gul. The characters that represent phonemes are grouped into syllable blocks. You can actually see this when you look at written Korean. Also, the fact that Japanese has three different writing systems gives it the flexibility of targeting different levels of the phonology. The fact is, syllables are comprised of phonemes, so it’s specious to suggest that these are totally discrete entities for writing. I have not read McGuinness’s book. Sounds to me like she is imposing a typology on writing systems, which is fine, but there’s always much to be learned by studying what resides at the boundaries in any typology. Most orthographic linguists would classify abjad systems (like Hebrew or Arabic) as alphabetical; indeed, they are the origins of all true modern alphabets. It’s also specious to peel apart the orthographic representation of phonological phenomena from the orthographic representation of meaning. Chinese is not purely syllabic; it’s logographic. Sumerian writing had different stages, including logographic, and eventually in cuneiform, pretty phonemic writing. I’m guessing based on the title of McGuinness’s book that she is making a pedagogical argument based on her linguistic typology, which is often where typologies go . . .

  9. shawnamay says:

    Thank you Gina! I am studying with Michel and have just ordered your products. I plan to take your Lexinars as well. I am a speech-language pathologist of 19 years. Nine of those years were spent teaching in higher ed. After reading this post I understand the difference between phones, phonemes, and allophones. I also understand hoe phonemes are distinctive and represent a change in meaning although they do not have meaning themselves. This was something I struggled to understand for years! I know meaning fit into the mix somewhere but was not sure how. I now see that when SLPs take phonetics we should learn phonetics, not a simplified version of phonetics. We only learned a few of the allophones and diacritical marks. They are all important! We learned nothing about phonemes or meaning. I recall trying to teach the difference between all of the ” ph” words to my students because I realized they had no idea of the differences. I did my best on my own but how I wish I could go back! Academia is certainly behind! I am so grateful for you and this community! I am pretty ticked off I am learning the truth at age 45.

    • Well my goodness Shawna for some reason I’m only just now seeing this, two months after you posted it. Sorry about that. I appreciate your comments and your clarion call as to the factors that are sorely lacking in most SLP programs (not to mention teacher education) when it comes to what phonology is — and isn’t. I hope you can join me in a LEXinar soon!

Leave a Reply

Your email address will not be published. Required fields are marked *