Portuguese Morphophonology: A Generative-Markedness Approach

Table Of Contents

• Cover
• Title
• Copyright
• About the author
• About the book
• This eBook can be cited
• Table of Contents
• Introduction
• 13 Regular Verbs: Inflection
• Introduction
• Transparent Regular Past Tense Verbs
• Opaque Past Tense: The Preterite
• Future Tense Regular Verbs
• The Present Tense
• The Present Subjunctive
• Gerunds
• ‘Past’ Participles
• Conclusions
• Appendix A Rewrite Rules
• Appendix B 2nd and 3rd Conjugation Root Vowel Metaphony
• Appendix C Morphophonemic and Rewrite Rules as They Appear in the Text
• Appendix D Rules in Their Order of Application and Sample Derivations
• Appendix E Rir, Crer, Ler: Irregular Present Tense, Regular Past Tense
• 14 Irregular Verbs: Present Tense
• Introduction
• Vocalism
• Consonantism
• Verbs of Aberrant Tonicity and Suppletive Verbs
• Conclusions
• Appendix A Rules as They First Appear
• Appendix B The Grammar of Irregular Present Tense Verbs
• Appendix C Suppletive Verbs: Rule Order and Derivations
• 15 Past Tense Irregularities
• Introduction
• The Imperfect Past Indicative
• The Preterite Indicative, Key to the Marked Past Irregular VerbsConclusions
• Appendix A Rules as They First Appear
• Appendix B The Grammar: Lexical Rules and Morphophonemic Rules in Their Order of Application
• 16 Verbal Derivation
• Introduction
• Part 1 Verbal Derivation with Theme Vowels
• Part 2 Intermediate Suffixes
• Part 3 Prefixes
• Conclusions
• Appendix A Rules in the Order They Appear in Chapter 16
• Appendix B
• Part 1 Rules in the Order They Apply in Verbal Derivation
• Part 2 Sample Derivations
• Appendix C Root Vowel ⇔ Nominal Vowel
• Appendix D Prefixes and Verb Formation
• Postscript 5 Verbal Hierarchies in the FrePOP Lexicon
• Restrospective Overview
• A Quixotic Endeavor
• References
• Index

←viii | ix→

Introduction

§Vi.0.0 Contrary to what obtains in Portuguese nominal inflection, where two morphological categories, gender ([±feminine]) and number ([±plural]), constitute the outer layer of word endings, verbal inflectional morphology is much more elaborate. Portuguese verbs are inflected with suffixes for grammatical person, tense, mood, and aspect. And there are six different grammatical persons (1^st, 2^nd, and 3^rd person singular and plural), three different tenses (past, present, future), two moods (subjunctive and indicative), and two past tense aspects (perfect and imperfect). The underlying structure of all finite verbs is:

Formula V¹i.²1

This means that all finite verbs begin with a stem that can be monomorphemic (i.e. a root). Like the nominals, all regular verbs have roots, but what turns roots or stems into verbs are deriving suffixes known as theme vowels (TV). These vowels derive three classes of verbs: 1^st, 2^nd, and 3^rd conjugations. After the theme vowels come the ←ix | x→tense/mood/aspect morphemes (TMA), most of which, like the theme vowels, can emerge word-finally, but the TMA morpheme is followed by person/number suffixes (P/N), some of which are null, some of which end verbs conjugated for every tense and mood.

§Vi.0.1 The goal of Chapters 13, 14, and 15 is to identify, analyze, regulate, and evaluate these verbal classes and the morphophonology (suffixes and rules) associated with their tenses, moods, aspects, and grammatical persons. Chapter 13 addresses regular verbs in all tenses, moods, and aspects. Chapter 14 examines present tense irregular verbs in both the indicative and subjunctive moods. And Chapter 15 explores the irregularities entailed in past tense verbs in their perfective and non-perfective aspects and in their subjunctive and indicative moods. Each chapter proposes a set of rules to account, to the extent possible, for the patterns entailed in the verbs’ conjugations. Chapter 16 addresses patterns of verb derivation.

§Vi.0.2 In Chapter 13, a set of rules accounts for the patterned formation of myriad verbs in all their tenses, moods, and aspects as well as their P/N inflections. Of course not all the rules apply to all the verbs since different phenomena occur in different classes of verbs. However, the regular (i.e. rule governed) phenomena occur in large classes of verbs. It is possible, thus, to define regular verbs as those verbs whose formal features can be described and generated by the set of rules in Chapter 13.

§Vi.0.3.0 Irregular verbs share many features with regular verbs, but to account for their varied forms in a morphophonological grammar, they need more than the set of rules for regular verbs. Thus, irregular verbs are those whose patterns require extra provisos and rules. Sometimes these patterns will obtain in several verbs, but some of the irregular verbs require rules that pertain only to them. Sometimes the past tense verbs will bear scant similarities to their present tense counterparts. Sometimes only a single aberrancy sets particular verbs apart from the regular verbs. And sometimes verbs are regular in the past tense but irregular in the present. But for any verb to be irregular, the set of rules that generate its forms must entail at least one special rule that is not among those that generate regular verbs. On the other hand, some irregular verb forms (e.g. perfective participles) are exempt from the rules that characterize regular verbs.

§Vi.0.3.1 Certain phenomena that obtain in particular verbal roots do not, in themselves, qualify as verbal irregularities. For example, the roots of verbs such as crer ‘to believe’, ter ‘to have’, sair ‘to leave’ end ←x | xi→in consonants: the root of crer is #kred+ (cf. credo ‘creed’); ter, #ten+ (cf. tênue ‘tenuous’); sair, #sal+ (cf. saliente ‘salient’). Loss of the /d+, /n+, and /l+ in these infinitives and in these verbs’ finite forms does not suffice to consider them irregular because these losses occur in nouns formed from the same roots: #kred+, crença ‘belief’; #ten+ teor ‘conceptual content’ and conteúdo ‘material content’; #sal+, saída ‘exit’. Likewise, the root vowel metaphony /e/ ≈ /ɛ/ in verbs such as meto [me.tŭ] ‘I insert’ ≈ mete [mɛ.tə] ‘3^rd inserts’, or the /u/ ≈ /ᴐ/ metaphony in verbs such as durmo [dur.mŭ] ‘I sleep’ ≈ dorme [dᴐr.mə] ‘3^rd sleeps’ does not make these or the many verbs like them irregular because given the proper (motivated) underlying forms, these metaphonic shifts are entirely predictable and rule governed.

§Vi.0.4.0 The Aurélio’s computerized list of the 3,000 most frequent words found in print during the 20^th century helps to put irregular verbs in perspective within the totality of Portuguese verbs. That list of 574³ verbs indicates that 19.13% of the most frequent words were some instantiation of those verbs—either finite forms or verboids and infinitives. Tables Vi.1a–b display data pertinent to this study:

Table Vi.1a Verb Class

Verbs In The Entire Aurélio (12,913)
	1st Conjugation			2nd Conjugationa			3rd Conjugation
Totals:	11,482	88.92%		753	5.83%		678	5.25%

a The verb pôr ‘to put’ and its related (e.g. opor ‘to oppose’) entries belong to the 2^nd conjugation.

Table Vi.1b Verb Class and Irregularity

574 Most Frequent Verbs
	1st Conjugation			2nd Conjugation			3rd Conjugation
Regular:	366	99.20%		83	72.81%		40	44.00%
Irregular:	3	0.80%		31	26.19%		51	56.00%
Totals:	369	64.28%		114	19.86%		91	15.85%

§Vi.0.4.1 These data confirm that both among the most frequent verbs and all the verbs listed in the Aurélio, the 1^st conjugation is the ←xi | xii→unmarked verb class because it is the most prevalent: 88.92% of all verbs, 64.28% of the most frequent verbs. Beyond this, within the most frequent verbs the 1^st conjugation member verbs are overwhelmingly regular: only three 1^st conjugation verbs are irregular. The 2^nd conjugation verbs are significantly less prevalent: 5.83% of all verbs, but 19.86% of the frequent verbs. They still evince a relatively high degree of regularity (72.81%) among the frequent verbs. The 3^rd conjugation is the least prevalent: 5.25% of all verb classes, and 15.85% of the frequent verbs. Among the most frequent verbs, those in the 3^rd conjugation are more irregular (56.00%) than regular (44.00%).

§Vi.0.4.2 The 2^nd and 3^rd conjugations’ scant difference in total numbers (753 or 5.83% of the verb database vs. 678 or 5.25% of that database) does not in itself compel one to consider the 3^rd conjugation more highly marked than the 2^nd. Both are minorities vis à vis the 1^st conjugation, and the 2^nd is only 1.11 times as numerous as the 3^rd conjugation. Conversely, the 1^st conjugation outnumbers the 2^nd and 3^rd conjugations combined by a ratio of 8.82/1. However, the 3^rd conjugation’s high degree of irregularity, at least among the most frequent verbs, suggests greater markedness. Of the 574 most frequent verbs, 85.19% are regular, which means that 14.81% of the verbs people are most likely to use in speech, to hear, and to write are irregular to greater or lesser extents.⁴

§Vi.0.4.3 Concerning verbal usage both in written and oral discourse, the database used in determining the frequency of Portuguese morphophonemes has been revealing. Table Vi.2 contains the numbers of regular and irregular verbs found in the texts that comprise this study’s database:

Table Vi.2 Verb Forms in the Database

	Regular Verbs		Irregular Verbs
Sources	Regular		Irregular Form	Regular Form	Infinitive	Total
Disney	65		55	16	12	83
Maria A	48		69	7	12	88
Meireles	53		36	12	7	55
Rodrigues	93		54	25	12	91
Leiria	43		36	15	3	54
Reis	44		16	3	3	22
Mira Mateus	37		17	2	2	21
Totals:	383		283	80	51	414
Percentages	48.05		35.51	10.04	6.40	51.94

←xii | xiii→§Vi.0.4.4 The database contains 797 verbs (including infinitives and past participles). Its texts vary from an aged villager’s personal narrative to literary analysis and an introduction to a phonological treatise. Over all, irregular verbs outnumber the regular ones by a factor of 1.06 irregular verbs for every regular verb.⁵ Beyond this, these texts suggest that one can expect slightly more than one third of verbs occurring in random texts to be irregular forms of irregular verbs. Some 58.09% of the verbs will occur in some sort of regular form. And, perhaps, some 6.40% of the verbs in representative texts will be infinitives, i.e. nominal lexical items.

§Vi.0.4.5 Concerning the texts themselves, one notices that in the less than formal texts (Disney, Maria Augusta, Meireles, Rodrigues, and Leiria), only Rodrigues’s text contained more regular (93) than irregular (91) verbs. Conversely, the academic texts (Reis and Mira Mateus) were weighted much more heavily toward regular verbs: Reis, ←xiii | xiv→two regulars to one irregular; Mira Mateus, 1.85 regulars to one irregular. This is because of the erudite nature of these texts and their specialized vocabulary (including often very academic 1^st conjugation verbs). On the other hand, Maria Augusta used a considerable number of repetitions and question tags (Não é ‘Isn’t that right?’), which raised her irregular verb tokens far beyond the regular ones (1.83/1). Table Vi.3 contains the actual verbs whose irregular forms occurred in the database:

Table Vi.3 Irregular Verb Tokens: Database

Verb	Gloss	Tokens		Verb	Gloss	Tokens
ser	to be	98		pôr	to put	4
ter	to have	36		dar	to give	2
estar	to be	31		caber	to fit	1
ir	to go	19		cobrir	to cover	1
vir	to come	19		crer	to believe	1
haver	to be	17		eleger	to elect	1
querer	to want	10		entregar	to deliver	1
fazer	to do	9		escrever	to write	1
ver	to see	9		expelir	to expel	1
dizer	to say	8		opor	to oppose	1
poder	to be able	6		ouvir	to hear	1
saber	to know	5

§Vi.0.4.6.0 These 23 verbs are responsible for 411 tokens in the database, but only 279 of them emerged in their irregular forms. Indeed, five of these verbs (cobrir, eleger, entregar, escrever, expelir) are irregular only in their past participles, which account for their presence in in this study. And, of all these items, only expelir was not among the 574 most common verbs. Thus, only 17 different verbs provided 274 of the 282 incidences of irregular forms—66.66% of all the verbal tokens in the Aurélio’s dataset.⁶←xiv | xv→

§Vi.0.4.6.1 Whereas Table Vi.2 can provide a working hypothesis concerning the prominence of irregular verbs in texts, and Table Vi.3 can provide hypotheses concerning relative frequency of these irregular verbs, a much larger database with even more varied speech and writing registers is necessary for establishing definitive notions of the roles of irregular verbs and irregular verb forms in the over all language. The FrePOP database actually provides a means for evaluating the relative frequency of irregular verbs such as those in Table Vi.3. Table Vi.4 demonstrates the relative frequency of the 23 verbs in Table Vi.3 plus one totally regular (frequent) verb, comer ‘to eat’:

Table Vi.4 Relative Frequency of 24 Verbs (293, 281 Tokens)^a

	Totals	%		Totals	%
ser	112,365	38.31	saber	5,781	1.97
ter	30,934	10.55	pôr	5,053	1.72
estar	25,959	8.85	comer	2,122	0.72
ir	20,225	6.90	ouvir	1,197	0.41
fazer	16,014	5.46	escrever	892	0.30
haver	15,619	5.33	crer	723	0.25
dizer	13,524	4.61	entregar	326	0.11
dar	9,534	3.25	caber	314	0.11
poder	9,174	3.13	cobrir	149	0.05
querer	8,700	2.97	opor	91	0.03
ver	8,128	2.77	eleger	45	0.02
vir	6,404	2.18	expelir	8	0.003
subtotals:	276,580	94.31		16,701	5.694
totals:	293,281

a M=12,254.92, SD= 22,823.51, Z=2.801, ser: outlier. It is worth noting that the first four most frequent irregular verbs are in the same order of frequency in both the FrePOP and the Markedness databases. Indeed 11 of the first 12 are the same.

The figures in Table Vi.4 represent the sums of these verbs’ infinitives and the finite forms of all their tenses and moods including their personal infinitives. The FrePOP data also indicate the relative frequency of these finite forms, including their overtly inflected infinitives:

Table Vi.5 Relative Frequency of Finite Verbs

TMA	Counta	%
Present Indicative	145,916	62.87
Imperfect Past	36,769	15.84
Preterite	25,978	11.19
Pesent Subjunctive	7,073	3.05
Fututure Indicative	4,857	2.10
Past Subjunctive	3,318	1.43
Personal Infinitive	2,546	1.10
Future Subjunctive	2,518	1.08
Conditional	2,297	1.00
Past Anterior	810	0.35
Total:	232,082	100.01

a M=23,208, SD=44,765.40, Z=2.29, outlier: present indicative, Z=2.74.

←xv | xvi→Using the frequency of occurrence criterion for assigning markedness, the data in Table Vi.5 indicate that of all tenses and moods the present indicative is the unmarked form, and the past anterior is the most highly marked. In the past tenses, the imperfect is the least marked, and the past anterior the most. Finally, the FrePOP data also supply an account of the person/number hierarchy:

Table Vi.6 Person/Number Hierarchy

Grammatical Person	Tokensa	%
3rd Singular	178,786	60.58
Bare Infinitive	34,626	11.73
3rd Plural	34,414	11.66
1st Singular	29,576	10.02
1st Plural	9,587	3.25
2nd Singular	6,050	2.05
2nd Plural	718	0.24
Totals:	295,123	99.53

a M=41,965.29, SD=61,946.06, Z=2.02, outlier 3^rd singular, Z=2.21.

It should come as no surprise that Table Vi.6 vouchsafes the 3^rd person singular forms as the unmarked person/number. Likewise, the rarity of the 2^nd person plural forms is not surprising. On the other hand, the uninflected infinitive edges out 3^rd person plural for the ←xvi | xvii→second most frequent form of verbs. This is to be expected since many paraphrastic constructions e.g. ele vai falar ‘he’s going to talk’, ela quer trabalhar ‘she wants to work’, eu queria comer ‘I wanted to eat’ contain infinitives.⁷ In every overtly inflected set the singular forms outrank the (more marked) plural forms: 3^rd S.>3^rd Pl., 1^st S.>1^st Pl., 2^nd S>2^nd Pl.

§Vi.1.0 Since each verb class has a theme vowel that is a deriving rather than inflecting suffix (§Vi.0.0), it is appropriate that these morphemes be evaluated here, in the introduction to the chapters dealing directly and exclusively with verbs. As Tables Vi.1a-b have suggested, the 1^st conjugation is the unmarked conjugation because it is far and away the most prevalent of the three verb classes, and 1^st conjugation verbs are overwhelmingly regular. Beyond this, the 1^st conjugation theme vowel (+a#) is the default verb deriver—almost all new verbal coinages are 1^st conjugation. Thus, the 2^nd and 3^rd conjugations and their theme vowels are the marked verbal cases—their relative rarity and low productivity make their information content greater than that of the 1^st conjugation verbs. The following matrix reflects their marked status:

Matrix Vi.1 Verb Conjugation Classes According to Theme Vowels

	+a#	+e#	+i+
2nd Conjugation	-	+	-
3rdConjugation	-	-	+

§Vi.1.1 For the purposes of this study, Portuguese verbs have two parameters or morphological marks that identify three distinct verb classes. The unmarked case stands in contrast with the other two cases since its specifications are the negation of the marks identifying the other two classes. And, beyond +i+’s infrequency and its restriction (it does not emerge word-finally),⁸ it has the highest percentage of ←xvii | xviii→irregular verbs among the most common verbs. Phonologically it is the most highly marked of the three theme vowels [+palatal, +high]—two positive specifications opposed to one for either +a# or +e#.

§Vi.2.0 As a final preview to the work on verbal inflection, Tables Vi.8 and Vi.9 contain the irregular verbs and their peculiarities examined in Chapters 14 (Irregular Present Tense) and 15 (Irregular Past Tense):

Table Vi.7 Verbs and Their Irregularities

	Present Tense							Past Tense
	1st P ➔ Subj.		Supletive		Vocalism
	1st P.	Subj.	Indic.	Subj.	TV➔Ø	RVM	Other	Imperfect	Marked
acudir						✓
agredir						✓
caber	✓	✓							✓
cair	✓
crer	✓
dar	✓								✓
dizer	✓	✓			✓				✓
estar	✓			✓					✓
fazer	✓	✓			✓				✓
haver	✓		✓	✓					✓
ir	✓	✓	✓						✓
ler	✓	✓
medir	✓	✓
odiar							✓
ouvir	✓	✓
pedir	✓	✓
perder	✓	✓
poder	✓	✓							✓
pôr	✓	✓						✓	✓
prazer	Ø				✓				✓
polir	✓	✓
produzir					✓
querer		✓			✓				✓
requerer	✓	✓
rir	✓	✓
saber	✓								✓
sair	✓	✓
ser	✓		✓	✓				✓	✓
ter	✓	✓			✓			✓	✓
trazer