Margo's Magical Letter Page
It has been a much discussed issue among linguists when we should consider a a string of sounds a suffix. I personally don't worry too much about whether a thing should be labeled a suffix or not. What I will discuss below is when a thing in fact is labeled a suffix. I've been particularly influenced in my thinking on this subject by the linguist, Dwight Bolinger. He points out that the notions 'suffix' and 'prefix' are much less well-defined than we would like to think. Given the data on phonosemantics I present in this page, it could be argued that every single vowel and consonant is some kind of prefix or suffix or root (i.e. some kind of morpheme).
What is really going on when we talk about suffixing or prefixing? There are at least two major processes happening. One is iconic or inherent meaning, in which the meaning is inherent in the form, and the other aspect of meaning is reference, in which a string of sounds is mapped onto something in the world.** (this is supposed to be a mark for a footnote, so see below...) Now suffixation is essentially the result of performing reference. The great English language says to herself, "One of the senses I'm going to allow for the stressless final syllable '-er' is 'a person who...'." That process has an element of arbitrariness to it, which is what deSaussure was talking about when he said that the mapping between sound and referent is arbitrary. A language by no means has to choose a stressless final '-er' syllable to actants (people who do things), and languages in general don't. It's not true, however, that the mapping is completely arbitrary. That sense 'person who' is much more likely to be applied to a stressless final '-er' than to a root 'splonk'. 'Person who' (being as it is inherently predicative of an action) is not the kind of thing that tends to be associated with roots, but rather with prefixes and suffixes and other little bobbles floating around in a language. Furthermore, /r/ is the most hyperactive of consonants, so it's not surprising that it should be associated with active people.
How do we in fact use the word 'suffix'? We call something a 'suffix' if it appears after the last root of a word, and if we are conscious of what referent it carries. We call the '-er' in 'player' a suffix, because we know consciously that it refers to 'a person who...'. The longer the string is, the more arbitrary its reference tends to be, and the more language-dependent it tends to be. That is, the strength and energy and reinforcing qualities of the consonant /s/ are quite common in languages all over the word, but the use of 'strangle' to refer to a violent action is much more narrowly limited to English.
There is another process that affects how we define suffixes. I will give an example of this process. Take the word 'editor'. The -or on the end is generally thought of as a suffix meaning 'person who...'. If you add the suffix '-ial' to produce the word 'editorial' meaning 'pertaining to editors', the '-or' still feels like a suffix, though perhaps a little less prominently than in 'editor' taken alone. If you then, however, take the whole word 'editorial' and use it to refer to something more narrow than the sum of the parts (edit, or, ial), that is, to 'a newspaper column', then the '-or' almost doesn't feel like a suffix at all any more. And the more frequently you repeat this - stick on suffixes and use the result to refer to something narrower than the sum of the parts, the further that -or suffix recedes into the subconscious (non-suffix status): editorialize, editorialization.
This is the story of the latinate morphology in English. So much of this is going on that you no longer have much of an idea at all what suffixes mean or are. We see '-ate' as a verbalizing suffix, as in amalgamate, fumigate, investigate. How about this '-ate' in 'editorializations, given that there's no 'editorializate'. That's a different '-ate' you say? Then what about 'investigations'. Is that '-ate' a different suffix from the '-ate' in 'investigate'? Is it different from the 'ate' in 'editorializations'? And you think this is a mess? You have no idea.
Check out 'antidisestablishmentarianism', which is quite falsely supposed to be the longest word in English. Do you have any idea what all those prefixes, suffixes and roots actually mean?
anti (sort of a negative thing, cool)
dis (another negative, still pretty cool)
ish (publish, furnish, etc. means what exactly?)
ment (like fragment, parchment, foment,...??? Hmmm. It means, well... it means...)
ar (dictionary, extraordinary, regular... you all see the relationship between these 'ar's as clear as day, right?)
i (okay, I would make it hard on you... we'll put the 'suffixes' 'i' and 'an' together)
ian (not so bad, a person associated with, or something like that)
ism (makes an abstract noun, piece of cake)
You see, the closer you get to the edges of the word, the more obvious it is what those affixes mean. But there is a logic even in those internal suffixes. That logic is not unrelated to the logic of sound. The weaker reference becomes as a factor, the stronger the influence of sound. I'm not yet prepared to present anything detailed and coherent on this, but it's another one of my hobbies.
**This mapping is restrained by semantic classes. For example, in English words with very concrete reference can quite generally be classified as animals, body parts, people, food, clothing, etc.. We don't assign special words to the class of white objects, or the class of smooth-textured objects. The kinds of subdivisions we make is already built into the language, and if you try to make up a word for something that falls outside these classes, like a word for white, smooth-textured objects, you do so at the peril of not being understood. Everyone is going to subconsciously expect your new word to fall into one of a very limited number of classes for the simple reason that every other word in English falls into one of these classes.
Margo's Magical Letter Page