Can the above words be divided further or are they monomorphemic words?

  • 1
    They can be divided as much as you want. Some people would divide them into syllables, or phonemes, or letters, or historical roots. What did you have in mind? "Morpheme" is a theoretical construct, so different theories might have different ideas about those words.
    – jlawler
    Commented Nov 4, 2022 at 15:21
  • Dyou mean Sammich?
    – vectory
    Commented Nov 4, 2022 at 16:30

2 Answers 2


It depends on your theory of morphemes and how to diagnose bimorphemicity. A fairly restrictive theory of morphology holds that a word can be decomposed into multiple morphemes only if the substrings of the word defining the morphemes have consistent underlying phonological shape, an identifiable meaning, and occur in at least two words. Under that theory, cranberry, gooseberry and strawberry are monomorphemic (because cran only appears in cranberry, and the meanings of "goose" and "straw" are unique to these particular words). It has been arguen for nearly half a century in generative grammar this theory is too restrictive.

In fact, cran exists in cranicot, crapapple, which incidentally supports the decomposition of apricot into two parts (I'm not sure where the division is). Sandwich could be decomposed into sand and wich, the latter appearing in manwich, cheesewich and hamwich (all are honest-to-gosh words). As for the requirement of identifiable meaning, many apparently bimorphemic verbs of English have parts without clearly-identifiable meanings, for example re-fer, con-fer, pre-fer, re-mit, per-mit, sub-mit... You can find -kin in pumpkin and gherkin (both are cucurbits).

The problem is that we don't have a well-motivated theory that tells us when we must and cannot break a unit into one of more "morphemes". In fact, in certain theories of word formation (e.g. Anderson's A-morphous Morphology), there are no "morphemes". Under the fairly weak criterion "parts can be extracted and identified in other words", those are not monomorphemic words. It just depends on what your criteria are for claiming morpheme status.


A morpheme is the "smallest meaningful lexical item", according to Wikipedia. In other words morphemes are the smallest individual components of words that still carry some meaning or usefulness for speakers. Even something like a suffix or prefix that changes the part of speech/how the word is used is a morpheme, like "ly" in happily.

In its present day usage "sandwich" is a stand alone word, not a combination of sand + wich. Because if it was, then the "sand" in sandwich would share some meaning with the word "sand" which it doesn't. And "wich" would have to have its own meaning/function as well, which it doesn't in this word.

However, according to the Online Etymology Dictionary the word sandwich is named after John Montagu the Earl of Sandwich, which is a town in England. If you go back to the Anglo-Saxon roots of the place name Sandwich - "Sandwice" in William the Conqueror's 1086 "Doomsday Book" surveying England and Wales - then you will find that it is in fact a combination sand + wic, a suffix meaning a dwelling where trade takes place. So in that context "Sandwich" certainly consisted of two separate morphemes, which combined to mean a "market town on sandy soil", according to Wikipedia.

But over the last 260 years since the term was first coined to refer to slices of meat between bread, "sandwich" has become a word consisting of a single morpheme that cannot be further broken down. If you are talking about the town of Sandwich, however that's another story.

I have not researched pumpkin and dictionary to the same degree. Based on my current understanding I would say that pumpkin is monomorphemic because it is not a combination of "pump" and "kin", it just happens to sound similar to those two words. Dictionary however I do think is a combination of at least two morphemes "diction" and the suffix "ary", meaning "thing belonging to or connected with" according to Merriam-Webster (think apiary or aviary). I say at least two, because I think "diction" could possibly also be broken down into "dic" and "tion".

  • While not necessarily generative (debatable), -wich is a clearly defined suffix seen in numerous other toponyms. Perhaps not everyone understands it, but many do.
    – cmw
    Commented Nov 4, 2022 at 16:51
  • 3
    Correct, wich is a well known suffix in place names (Greenwich, Norwich, etc). But does it actually function as a suffix in the word "sandwich"? I would say no. If you are talking about the town of Sandwich in Kent, then I would say yes. Commented Nov 4, 2022 at 17:25
  • 1
    This is tricky then, because a sandwich is named after Sandwich. I'm not sure how exactly that plays a role, but it feels odd to exclude it.
    – cmw
    Commented Nov 4, 2022 at 18:14
  • 3
    @cmw Considering that a morpheme is the smallest meaningful, lexical item, we can only speak of morphemes for a given unique language – that is, for a given speaker. The vast majority of English speakers, I would venture, are unaware that the noun sandwich is named after a place, and to them, it is definitely monomorphemic. Plus, dissociation from names happens. Would you consider cheeseburger to contain the ‘morpheme’ Hamburg (itself actually dimorphemic)? Commented Nov 4, 2022 at 23:18
  • Also, to me pumpkin is definitely dimorphemic: pump is a cranberry morpheme that refers vaguely to gourds, while -kin is a diminutive suffix. Commented Nov 4, 2022 at 23:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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