Consider the following examples:

Fact -> /fækt/

Hard -> /hɑːrd/

Paint -> /peɪnt/

In all these words, the clusters in the coda are easy to pronounce. However, when these clusters come in the onset of a word, they become hard to pronounce.

/rd/, /nt/, /kt/ etc do not occur in onsets.

There are some clusters that are easy to pronounce in the onsets but difficult to pronounce in the codas.

Examples: Plant -> /plɑːnt/ -- /pl/ is easy to pronounce in the onset but difficult to pronounce in the coda. There are many others.

So why are some onset clusters difficult to pronounce in the codas and some coda clusters are difficult to pronounce in the onsets? Can anyone offer some insight?

  • 1
    Look up the sonority heirarchy. – curiousdannii Jun 7 at 14:03
  • @curiousdannii, I've already read that but didn't understand how it was related to sonority heirarchy. Could you explain it in an answer? – Joyful Sadness Jun 7 at 14:08
  • Note that your question qualifies rather unsuitable for this SE. First, it's a language specific question, it's about English. Also, it's opinion based, since it's only for you that it's hard to pronounce some consonantal clusters, for me it's very easy. Also note that "hard" is very subjective, for example in Russian kto "who" is an extremely frequent word. – Yellow Sky Jun 7 at 19:22

English has very strict rules as for which syllable-initial consonant clusters are allowed and which are not. If you are a native speaker of English and don't speak any other language, this means that for all of your life you have been beginning syllables only with the clusters that English allows, you just had no experience to start syllables in any different way. With some training one can pronounce practically any combination of consonants as a syllable-initial cluster.

English builds syllable-initial biconsonantal clusters (2 consonants) according to the two allowed models:

  1. non-sonorant + sonorant: after /p/, /t/, /k/, /b/, /d/, /g/, /f/, /θ/ there follows /l/, /r/, or /w/, if in the resulting pair the places of articulation of each consonant are different. Thus, /w/ being a labial never follows labials (/p/, /b/, /f/), and coronal /l/ cannot follow another coronal consonant (/t/, /d/, /θ/);

  2. /s/ + occlusive: the fricative obstruent /s/ can be followed by an occlusive /m/, /n/, /p/, /t/, or /k/.

Also, /s/ and /ʃ/ take part in the clusters of the first type: /s/ can be followed by /l/ or /w/; and /r/ can follow /ʃ/.

Clusters of three consonants are constructed by combining the two aforementioned models: sp-, st-, sk-, created after model 2, are followed by the sonorants which are allowed after /p/, /t/, and /k/ according to model 1.

One third of all the English consonant phonemes, eight, cannot be a part of a syllable-initial consonant clusters. They are /v/, /ð/, /z/, /ʒ/, /tʃ/, /dʒ/, /h/, /ŋ/.

/j/ has a special status. Any vowel can follow /j/ when /j/ is a single word-initial consonant. But if it follows another consonant, only /u/ can follow such /j/, pronouncing other vowels after a non-initial /j/ is not typical. On the other hand, the choice of the preceding consonant for /j/ is practically unlimited, /j/ can follow consonants /v/ and /h/ that cannot be a part of any other cluster, it can follow /l/, /m/, and /n/ that otherwise cannot begin a cluster. All this pushes one to treat /ju/ as a diphthong, a vowel, but that's quite a different story.

What we can see is that in the English syllable-initial clusters non-sonorants combine with sonorants, but two sonorants cannot combine, non-occlusives combine with occlusives, but two occlusives cannot combine either. In your examples, it was "hard" to pronounce those clusters as initial since /kt/ is "two occlusives" (forbidden), in /rd/ and /nt/ a sonorant precedes a non-sonorant, the order which is forbidden, too.

| improve this answer | |

"Hard to pronounce" is a popular cover term regarding phonological patterns the violate the rules of the language, and it doesn't have anything to do with overly-taxing physical movements. In this case, there is a concept of "sonority", which is a pairwise relationship between sounds and how they are ordered within the syllable, that sonority should rise then fall. There are many proposed sonority subdivisions, but one popular division is obstruent < nasal < liquid < glide < vowel, which means that syllables can start with a segments on the left end of the scale, be followed by segments further to the right, then the "direction" reverses with the vowel and syllables can end with sequences of segments that increase in sonority.

Conventionally, sonority is assigned a number from 1 to something small like 7. The best attempts to assign a phonetic interpretation to "sonority" is that it pertains to the opennesss of the vocal tract, which relates to dampening of vocal tract resonances.

This scheme doesn't explain everything about the ordering and possibility of consonant sequences in the syllable, for example doesn't explain why *tl is excluded from onsets when pl and tr are allowed. An alternative theory forgoes the concept "sonority" and explains the patterns in terms of perceptibility factors, e.g. the facts of r and d that make it difficult to hear the r in *rdæg and *gɑdr but not dræg and gɑrd.

| improve this answer | |
  • Was *gɑrd supposed to be *gɑdr? (Also, why should a sonority scheme explain why a specific cluster happens to be absent from a specific language? There are lots of sonorically acceptable clusters that happen not to exist in various languages, and tl- is common enough cross-linguistically.) – Janus Bahs Jacquet Jun 7 at 17:17
  • It shouldn't. If you accept that there are other factors determining segment-sequencing, then you don't need sonority, you can just refer to those other perceptual factors directly. – user6726 Jun 7 at 17:33
  • My answer here explains why *tl is excluded from onsets when pl and tr are allowed. :P – Yellow Sky Jun 7 at 21:09

1. /rt/, /nt/

For the second and third example, I'd argue that the sonorants /r n/ are syllabic. These can in principle not appear at the beginning of a word--in English--if they are perceived as part of the nucleus of the syllable, so we transcribe e.g. Rigveda, and vriddi, or add diacrits. Likewise, with a voiced onset there's no problem:


arguably the onset can vary widely, and there should be no problem to reduce it down to a glottalic nasal, a mere opening glottal plosive


or a slow rising nasal (?)


This indistinguishable from a vowel onset, insofar both are allomorph in English!

Whereas an unvoiced nasal stop doesn't even exist, I guess, if there are no other marked features.

It's curious that you didn't mention Pterodactyl. Arguably it is not too dificult.

2. /kt/, /pl/

I would have problems with the first example, /kt-/, too, because I'm not used to any reasonable realization, and any approximation would insert voicing or at least aspiration, because my /k/ is either an approximant or a fricative (typical for Low German, I guess, with traditional /k/ > /h/, etc).

Vice versa, obstruent + liquid would always be realized with voice, so we write appl-e. The notation is simply misleading. category could be reasonably transcribed /kt-/ for certain renditions, but we still hear /kat-/ (modulo correct vowel), so this is subject to cognitive linguistics. Therefore it's not even debatale whether the inclined bilingual speaker who makes a phonemic distinction would mark the difference, if we don't intend to. I'm pretty sure it's not even close as an approximation, so that's not debatable either.

3. /h/ and the case of Ru. kto

Vice versa, Russians typically can't pronnounce /h/ and substitute a fricative /x/ instead.

kto reflects *kъto (so in Old Church Slavonic), from PIE *kwis and then some, wherein ъ "denoted an ultra-short or reduced middle rounded vowel" that is now called back yer. Today in Russian, the letter marks hard vowels that are not palatilized, but historically it could cause palatization, hence the word reflects as e.g. xto (variously), or some notable voicing in kdo (Slovenien). If it caused some form of feature in Russian that is just not phonemic in Russian it will consequently not be transcribed (therefore the specific feature doesn't matter here), much less if it's ultra short.

For some annecdotal evidence:

Perhaps the most common objection against a proposed reconstruction which is raised time and again on general grounds, is that a linguistic form is impossible because it does not conform to typological expectations. The classic example is Brugmann’s reconstruction of nasalis sonans in 1876, e.g. in the first syllable of *kmtóm ‘hundred’. Brugmann published his article in a journal of which Curtius had made him co-editor before going on a journey. When the latter read the article after his return, he became so enraged that he dissolved the journal and started a new one, without Brugmann (cf. Pedersen 1962: 293). The new reconstruction has now been part of the communis opinio for over a century. The case of the nasalis sonans is particularly instructive because the new the- ory soon gained general acceptance. The same cannot be said of the hypothesis that the Indo-European proto-language had no more than a single vowel. It is therefore important to compare the two cases in order to establish the reason for the different treatment. Note that I am not primarily concerned with the correct- ness of the reconstructions but with their reception by the scholarly community. If we can find out what motivates our colleagues to agree or to disagree, it may be possible to save a lot of time when trying to convince them. There are two types of objection against the reconstruction of a single vowel for Proto-Indo-European. On the one hand, it is claimed that not all of the material can be explained from such a reconstruction. On the other hand, it is argued that there can be no such thing as a language with no more than one vowel. Both ar- guments have their counterparts in the reconstruction of the Proto-Indo-European syllabic resonants. In the case of the nasalis sonans, there were two factors which rendered the new reconstruction more palatable. While the concept of syllabic nasal was an innovation, the syllabic liquids l and r were familiar from Czech and Sanskrit. The new theory did not therefore affect the idea of syllabicity as a vocalic prop-

erty but only its distribution. Moreover, the class of possible reconstructed forms was not greatly affected because Brugmann recognized, beside the zero grade vo- calism of the syllabic resonants, a reduced grade vocalism which could be in- voked for those instances where others might see counter-evidence. It can be ar- gued that the real victory of the Sonantentheorie was eventually achieved by the elimination of the reduced grade. That was a development which took much longer than the acceptance of the nasalis sonans.


NB: *k'mt- > **k't- was wholy rejected. (I'm sure that has something to do with the topic, but please don't ask.)

| improve this answer | |
  • 2
    Will you, please, explain how /r n/ are sibilants? – Yellow Sky Jun 7 at 22:00
  • @Yellow-Sky instead of saying something usefull i rather chose a term i was faintly aware must be wrong, perhaps in the hope that it will become clear in context. please tell me, could I have meant syllabic? – vectory Jun 7 at 22:44
  • 2
    Maybe, 'syllabic' sounds more like what you explain there. But note, in different languages of the world, words can pretty well begin with syllabic consonants, for example, the name of the Indian book 'Rigveda' is pronounced as [ṛg.ve.ˈda] in Sanskrit, in Swahili the word for 'person' is mtu [ˈm̩.tu]. – Yellow Sky Jun 7 at 23:02
  • @yellow-sky, RFC to the added paragraph about kto, plz – vectory Jun 7 at 23:15
  • Both Slovene and Czech kdo are in fact pronounced [gdo], it's obvious that such pronunciation appeared by analogy of the word gde 'where' which in its turn appeared from *kъde after [d] had assimilated [k] by voicedness. The Russian хто is by Ukrainian influence, where kto > xto, gde > de. – Yellow Sky Jun 7 at 23:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.