Linguistics An Essential Introduction

As we said in Section 5.10, morphology is the study of words and their internal structure — since simple words do not have internal structure, this means in particular complex words. We have seen a range of such words and word forms now, most importantly, words consisting of a base and one or more derivational affixes, like /ˈlɑː.ləs/ lawless or /ʌnˈlɑː.fəl/ unlawful, words consisting of a base and an inflectional morpheme, like /lɑːz/ laws, and words consisting of two (or more) free morphemes, like /ˈlɑː.mən/ lawman.

In order to account for the internal structure of these words, we need to describe the bound morphemes along three dimensions: their form, their meaning and their combinatorial properties, i.e., (a) where they attach to the base (whether they are prefixes, suffixes, circumfixes or infixes), (b) what wordclass the base must have, and (c) what wordclass the derivative word has.

In morphology, such a description typically takes the form of a word-formation rule (WFR) — essentially, a rule for attaching a bound morpheme to a base. There is no generally agreed format for such rules, but all formats you are likely to encounter will make reference to the three dimensions just mentioned. A relatively typical example of such a format would look like this:

The first line shows the form: The base is represented by an X with a subscript specifying the word class, the affix is shown in the position it has relative to the stem. The combination of base and affix is enclosed in square brackets with a subscript specifying the word class of the derivative. The second line shows the meaning in the same way we have used throughout the book so far: either as a paraphrase in single quotation marks, or in uppercase. This notation can be read as follows:

The third line describes any additional conditions that must hold in order for the rule to apply (if there are such conditions).

The WFR for the affix {‑less} would look like this:

Note that we do not need a hyphen in front of the affix, as the order of base and affix is directly represented in the rule, but we will see below that there are good reasons to mark the boundary between base and affix anyway.

The word-formation rules for derivational and inflectional affixes follow the same structure — any differences between the two types of affix can be included among the conditions, if necessary. For example, the WFR for the plural affix {-z} would look like this:

Note that we need to specify that the base must be a singular noun — if we did not, the rule could apply recursively, adding plural affixes to words that already contain such an affix: *law-s-s-s-s-s-….

These WFRs may look simple, but they are very powerful tools for capturing the structure of polysyllabic words. Take the word unlawful. Recall that we pointed out that in order to derive this word, we must first derive the word lawful and then add the affix {un‑}. The other order would not work, because the word *unlaw does not exist. These facts follow automatically from the following WFRs:

The form *unlaw cannot be derived, because the WFR for {un‑} specifies that the base must be an adjective. Thus, the rule for {‑ful} (or another rule deriving adjectives from nouns) must apply first, giving us first lawful and then unlawful.

The two rules also illustrate the kind of conditions that we may have to specify for a WFR. The rule for {‑ful} specifies that the base must be an abstract noun — this is to avoid the formation of words like *homeful, *stainful, *breathful, *priceful, *windowful etc. Note that {‑less} does not have this restriction, so we get words like homeless, stainless, breathless, priceless, windowless.

The rule for {un‑} specifies that the base must refer to a “default property” — by this we mean properties that are assumed to be the “normal” case. We expect lawfulness to be normal state of affairs, so we get unlawful, but not *unlawless, we expect happiness to the the normal state of affairs, so we get unhappy, but not *unsad, we expect necessity to be the normal state of affairs so we get unnecessary but not *unsuperfluous, and so on. Admittedly, this concept is somewhat vague, but that is why linguistics is still an area of active research!

When you look at words created by a specific WFR, you will often note that they mean something more specific or even something slightly different than what the rule would lead you to expect. For example, the rule for {‑ful} states that the meaning of derivatives with this affix should be ‘having/causing X’, and for many cases, this is a decent definition: beautiful ‘having beauty’, successful ‘having success’, powerful ‘having power’, painful ‘causing pain’, stressful ‘causing stress’ etc. The first question would be how we know when the affix means ‘having X’ and when it means ‘causing X’. Why do we use beautiful for something that has beauty (a beautiful person), but not for something that causes beauty (*a beautiful surgery for a plastic surgery), and painful for something that causes pain (a painful surgery) but not something that has pain (a painful person for someone who is in pain)? The answer is that the WFR specifies the range of meanings that the derivate can have, but once the derivate is established, it will typically have a more specific meaning rather than the whole range of meanings. It may even develop a slightly different meaning altogether, like plentiful, which means neither ‘having plenty’ nor ‘causing plenty’, but ‘being available in great quantity’, or mindful, which means neither ‘having a mind’ nor ‘causing a mind’, but ‘paying careful attention to something.