Challenging Lexical Morphology

Table of contents

Abstract

1 Introduction

2 Word formation processes and lexical strata
2.1 Introducing affixes
2.2 Introducing lexical strata
2.3 Prefixes
2.4 Assignments oflexical strata
2.4.1 Productivity and irregular inflection at stratum 1
2.4.2 Regular inflection at stratum 2
2.5 The order of strata to apply

3 The refinement of the lexical phonology and morphology model
3.1 Blocking, lexical integrity and the Elsewhere Condition
3.2 Compounding
3.3 The order of processes within strata
3.4 The amount of required strata
3.4.1 The most prominent approaches
3.4.2 The Strict Cycle Condition
3.4.3 Step-by-stepconfrontation

4 Yet unsolvable problems within the model
4.1 "Multi-stratum" affixes
4.2 Lexical strata and bracketing paradoxes

5 The counter approach: roots as determinants for lexical strata

6 Conclusion

7 Bibliography

"Eigenständigkeitserklärung"

Abstract

In this term paper I examine several approaches to the bundling of both morphological and phonological rules, which is commonly referred to as stratification and which is the basis for the lexical morphology and phonology model. Therefore I first introduce the idea of strata with respect to their respective tasks and the order which they usually appear in (section 2). In the following I show up the first (minor) challenges within the model that eventually help to refine it (section 3 ). Furthermore I display problems that have not yet been solved, meaning inconsistencies within the stratification processes as they have been introduced (section 4). Right before the conclusion I briefly introduce a different approach (by Goldsmith 1990), which contradicts most of the remarks in the previous sections, but can eventually be refuted quite easily (section 5). In the conclusion I point out that the lexical morphology and phonology model has its problems (like probably nearly any other linguistic theory), but is most likely to be the most appropriate model for this motivation.

1.Introduction

It is probably an unsolvable task to find out how many words there actually are in the English language. However, we can be sure that, regarding the many varieties of English and all their word forms and loan words, we look at a number in the range of hundreds of thousands of words. Obviously most of them are strongly related to a variety of several other words, and the human brain's ability to crossreference new information with knowledge that has already been stored enables us in many situations to understand even words which we have never heard before. Someone with a concept of the English language will immediately understand the meaning of an adverb he hears or reads for the first time, even if he so far had never encountered this very word and had only known the respective adj ective. The simple suffixation of -ly (or /In/) will give him all the information he needs to understand the word perfectly in its context and its affiliated meaning.

Affixes in English generally have the power to change a word's meaning (although a close relation to the root word is usually preserved), its number or even its lexical category. For instance, adding the suffix -al to the noun nation results in the formation of the adj ective national. Considering this adj ective as the new root, a wide variety of suffixes can again be added in order to form new words: these can be nouns like nationalist, nationalism or nationality, verbs like nationalize or an adverb like nationally. All these suffixes (-ist, -ism, -ity, -ize, -ly) give out well-formed English words when attached to the adj ective national, but deliver ungrammatical results when attached to the noun nation (*nationist, *nationism, *nationity,...). Also, some of them preserve the root's stress pattern ( 'national ^ 'nationalist), others move the stress to another syllable ('national ^ nationality). In other cases affixes have even the power to delete, add or change sounds in the root.

In this term paper, I will examine several approaches to the distinction of morphological and phonological rules based on affixes regarding their functions and impacts, focusing on the problems which occur within these various models and analyses. I will start by looking at several attempts to find proper categories for affixes and their respective modes of operation. Secondly I will deal with a problem which all approaches share: none of them works without exceptions. In another two sections, the categorization models are confronted with phonological and morphological rules to examine whether they can co-exist in general. Some of them seem to have been solved within the models that various linguists have proposed, others remain problematic. In my conclusion, I will revise lexical strata in the lexical morphology and phonology model with respect to the different approaches in order to evaluate their functionality for the analysis of word formation processes.

2.Word formation processes and lexical strata

In this section I will briefly introduce the different kinds of affixes and a popular way for their categorization: lexical strata. Yet unanswered will remain the question if this categorization is always accurate and how many categories (= strata, levels, classes,...) are needed for a proper investigation. Instead, this section will deal with the interference of morphological and phonological rules that are both required to properly study word formation processes

2.1 Introducing affixes

The model of lexical phonology and morphology combines the analysis of morphological structures of words with its respective phonological rules in order to deliver a proper explanation of the development from one word to another. Therefore linguists act on the assumption that speakers of a language know the words of this language along with their internal structure and can intuitively process word formations by affixation or compounding (cf. Scalise 1986: 24). This involves knowledge about which word formation processes are licit in their language and which are ungrammatical:

(1)

Abbildung in dieser Leseprobe nicht enthalten

According to (1) it can be assumed that the formation of new words with the use of affixes takes place in a specific order, i.e. the lexicon is organized hierarchically (cf. Katamba/Stonham 2006: 89). Also, some affixes may apparently be attached to roots while thers can not.

(2)

Abbildung in dieser Leseprobe nicht enthalten

Examples (2a-b) show adj ectives derived from nouns^{1 2}, using the affixes -ful or -less. In these examples, the suffixation does not make any changes to the respective roots, meaning that both the sound and the stress pattern are preserved, even though the word is prolonged. (2b) shows that suffixation may cause changes in spelling, however these do not change the respective pronunciation. (2c) displays an example of a noun derived from an adj ective with use of the affix -ity ³, causing the root's stress to shift to another syllable. (2d-e) show that some affixes may even change sound changes inside of the root.

2.2 Lexical strata

Regarding the observations in the previous section, there seem to be two different types of affix in English (cf. Bauer 2003: 174). Affixes as shown in (2c-e) are non-neutral in a way that they tend to change the root which they are attached to. The affixes in (2a-b) are neutral, as they preserve the root. Chomsky and Halle refer to the behavior of non-neutral affixes as a weak boundary (between root and affix) and apply a strong boundary to the intervention of a root and a neutral affix (1968: 66ff.). Affixes also tend to differ according to their origins, as neutral ones seem to be Germanic, while non-neutral affixes are predominantly of foreign (mostly Latinate) origin (cf. Plag 2003: 168; Katamba/Stonham 2006: 91). In lexical morphology and phonology, these categories of affixes are usually referred to as strata, classes or levels. While there are various approaches to lexical strata, I will for the moment limit my remarks to a two-stratum model, as described by Katamba and Stonham (2006: 92f.), McMahon (2000: 36f.) and many more. Other models will be dealt with in the section 3.4.

Non-neutral affixes, like -ity and -ion, are thus referred to as stratum 1 affixes, including all affixes that cause changes in the stress pattern or the sound of the word. These include also affixes that let the stress fall on themselves ⁴:

(3)

Abbildung in dieser Leseprobe nicht enthalten

Coherently, neutral affixes are called stratum 2 affixes. They are attached to a root without influencing it, meaning that stress and sounds are preserved. The fact that the spelling of the word might actually change (as in (2b)) is irrelevant for phonological investigations.

2.3 Prefixes

Although suffixes have so far been the only center of attention, all these lexical rules apply to prefixes equally ⁵:

(4)

Abbildung in dieser Leseprobe nicht enthalten

2.4 Assignments oflexical strata

So far it has been worked out that affixes and the word formation processes they trigger can be categorized in two strata. In this section I will further examine the different phenomena at the strata to expose their respective qualities.

2.4.1 Productivity and irregular inflection at stratum 1

Generally speaking, stratum 1 affixes are more productive than stratum 2 affixes (cf. Fuster 2011: 77), meaning that they usually have the quality to change a word's lexical category when being attached to it:

(5)

Abbildung in dieser Leseprobe nicht enthalten

However, Stratum 1 affixes do not necessarily have to change the lexical category. Not only are there affixes at stratum 1 that preserve the category, but only irregular inflection (usually considered less productive) takes place at stratum 1:

(6)

Abbildung in dieser Leseprobe nicht enthalten

The examples in (6) cause some problems considering what has been stated about stratum 1 affixes so far, since in some cases no affix is at first sight observable at all. The affix -en looks like a stratum 2 affix in (6a), since no changes are made in the root. Nevertheless, (6b-c) deliver evidence for the fact that -en - used as an affix for the pluralization of nouns - triggers changes in the root's sounds and is therefore definitely a stratum 1 affix (if we follow the claim that all affixes belong uniquely toj ust one stratum; cf. section 4.1).

Yet harder to comprehend are the processes which take place in (6d-e). Ablaut and umlaut processes are not anymore used in modern English, since they have been replaced by the suffixation of -ed (for the formation of past tense verbs) or -sl-es (for the formation of plural nouns) respectively. Thus examples like (6d-e) have to be considered as hangovers from an earlier stage of the English language and its Germanic origins. However, since the stratum model is supposed to cover word formation processes in general, (6d-e) have to be assigned to stratum 1 due to the fact that the inflection changes the root, even though the affixes that might once have been used in this cases are no longer observable. (6f), finally, displays an inflection caused by the rules of word formation processes in Latin, which has been unvariedly transfered to English (cf. KatambalStonham 2006: 101ff.).

2.4.2 Regular inflection at stratum 2

As displayed in the sections 2.2 and 2.3, stratum 2 affixes are attached to roots without causing it to change in any way, i.e. both their stress patterns and their sounds remain the same. Like many non-neutral affixes, some stratum 2 affixes are also able to change a word's lexical category when attached to it:

(7)

Abbildung in dieser Leseprobe nicht enthalten

[...]

¹ This phonetic representation is (as will all phonetic representations of English words or morphemes in this term paper be) based on Standard American English. However, the crucial aspect in these cases (stress shift and/or vowel modifications) apply equally to other varieties of English, including Standard British English. I will add a representation in British English if necessary.

² The examples in (2a) might also be verbs, which is however not relevant in these cases.

³ It might be worth mentioning that the word original itself has already undergone affix-related derivation (origin -> original). This does however not have any impact on its own derivation, as is shown in section 3.3.

⁴ While this specialty is rather rare in English, it is quite common in Romance languages (cf. Saciuk 1969: 56).

⁵ The fact that prefixes are hardly mentioned in literature might be caused by their inconvenience. Many prefixes can be both neutral and non-neutral, depending on the respective root they are attached to. This problem will be a topic in section 4.1.