In ch. 3.2 I defined two parsing principles, Mother Node Construction (MNC) and Immediate Constituent Attachment (ICA). The former is a construction principle: it defines a method for recognizing abstract phrasal groupings on the basis of terminal elements within a parse string. Without such a principle, constituent structure would not be recognizable on-line, there would be no such thing as a Constituent Recognition Domain (CRD), and so on. Clearly, the construction of abstract syntactic nodes, and the correct assignment of words to their appropriate dominating nodes, is a vital prerequisite for the everyday use of human language.

This chapter will consider the theory of node construction in more detail. MNC is an extremely general and all-pervasive principle which has been largely sufficient to illustrate the wide range of ElC-derived word order predictions presented in chs. 4 and 5. But the construction of a higher node by an immediate daughter is not the only way of adding non-terminal structure to the terminal categories of the parse string, and I have already alluded to the existence of some additional construction principles in previous chapters. The purpose of this chapter is to define these principles and to consider the predictions they make for variation in both category selection and ordering across languages. A further goal is to consider how grammars have responded to the parser's need to recognize higher constituent structure in a rapid and efficient manner.

This study had its origins in my doctoral dissertation on Early Middle English changes of vowel quantity. Since I was a student at Vienna University, it is no big surprise that my thesis took the great Viennese philologist Karl Luick's treatment of the topic as a starting point. In some way, even, it started out as an attempt to translate the story Luick had told of the changes into the language of historical linguists of our time. As often happens with translations, however, mine turned out to become an interpretation, a deconstruction and eventually an almost complete recreation of the text it set out merely to ‘make understandable’. The obvious reason for this was, of course, that many of the Neogrammarian concepts Luick had employed have come to be refuted by the linguistic community and that even the very existence of sound laws that had long counted as well established has come to be questioned in brilliant and convincing ways by modern historical linguists. Most eye-opening to me, in this respect, was Donka Minkova's radical re-interpretation of Middle English Open Syllable Lengthening in her 1982 paper in Folia Linguistica Historica. At some stages during my work, then, I thought that the purpose of my study was to discover ‘errors’ in the stories of Luick and my other predecessors, to set them right and to make their accounts more ‘true’. Certainly, while I was working on my dissertation, such a heroic search for ‘truth’ appeared as a noble and worthy task to me and motivated me greatly.

Introduction

As follows from what has been said so far, this chapter will not simply describe the diachronic correspondences normally referred to as lengthening of vowels before homorganic consonant clusters. Rather, it will focus on their relation to the lengthenings discussed in chapter 1, and pursue the question whether HOL and OSL can be regarded as one great quantity change or whether they represented essentially different phenomena. If the data normally accounted for through HOL turn out to be compatible with the generalizations inherent to formula (1) of chapter 5, this will be taken to mean that there was indeed only one change. Should they turn out incompatible, however, it will have to be accepted that OSL and HOL were separate sound changes in their own right and that no unified account is possible. The hypothesis that formula (1) of chapter 5 ‘predicts’ all Early Middle English quantity changes will then be regarded as falsified.

The description of HOL will be carried out along the same lines as that of OSL. With the help of the Oxford Etymological Dictionary a list of more than 200 potential inputs that have survived into Modern English has been drawn up. It gives a representative view of the long-term implementation (that is to say, the constraints) of the process in question (see appendix II). The list has then been subjected to a similar kind of analysis as the Minkova corpus.

In its widest interpretation, Homorganic Lengthening was a process by which short vowels were lengthened, if they were followed by clusters of two consonants.

The standard descriptions

It is widely acknowledged among historical linguists that between the ninth and the thirteenth centuries English stressed vowels under went widespread quantity changes. The established way of describing these alterations is in terms of four distinct sound changes. The first of them, which has become known as Homorganic Lengthening (from now on HOL), seems to have made short vowels long, if they were followed by certain consonant groups (namely: mb, nd, ng, ld, rd, rs, rn, rð) – unless those groups were themselves followed by a consonant. It turned bindan into bīndan, cild into cīld, or climban into clīmban, to give a few examples. The second change is supposed to have made long vowels short, if they were followed by a group of two consonants. (Homorganic groups did not trigger the change, however; nor did groups that occurred at the beginning of words, such as: pl, pr, cl or tr.) This change, which is called Shortening before Consonant Clusters (SHOCC), is taken to have been behind such changes as that of kēpte into kepte, dūst into dust or fīfta into fifta. By the third change, then, long vowels are supposed to have been shortened if they occurred in the antepenultimate syllables of wordforms: suþerne and erende are thus said to have replaced sūþerne and ērende, for example. This change is commonly called Trisyllabic Shortening (TRISH). The fourth process, finally, is believed to have lengthened short vowels, if no consonant followed them within the same syllable. It is said to have turned maken into māken, weven into wēven or hopen into hōpen and is known as Open Syllable Lengthening (OSL).

As I have already stated, Modern English data show that OSL did not affect the vowels of all its potential targets equally. Rather, some type of OSL targets were more likely to be changed by the process than others. This is, basically, what statements such as Low vowels seem to have been more likely to be lengthened than non-low vowels express.

What I would like to highlight here, then, is that these statements have so far been taken to be relevant exclusively with regard to potential OSL candidates. The corpus from which they were derived contained no other items. On second thoughts, however, hardly anything seems to speak against widening their potential applicability and interpreting them as general tendencies behind Early Middle English vowel lengthenings. This point is crucial and I would like to elaborate the argumentation on which it rests, although it might be viewed as rather pedestrian in essence.

Take, once more, the statement about vowel height and imagine, for a momentary excursion, that there were no other ‘sound changes’ apart from OSL, by which vowels were lengthened. This means that the probability of vowel lengthening would have been zero among all non-OSL candidates, no matter what the heights of their vowels were. Therefore, if OSL candidates and other words were combined into a single set, the relation between lengthened low vowels and lengthened high vowels would basically remain the same, although the difference will be relatively smaller.