Branching lexical plural into greater and paucal

Abstract

Lexical plural merges in or below the categorizing head n⁰. It often yields abundance in quantity, as in Greek and Innu-aimun. However, it is reported that it denotes paucity in quantity in one language, namely Telugu, to the best of the researcher’s knowledge. The current study reports that Jordanian Arabic has lexical plural that is semantically similar to the lexical plural in Telugu. Specifically, it yields paucity in quantity. From a typological perspective, this study calls for branching lexical plural into lexical greater as in Greek and Innu-aimun and lexical paucal as in Jordanian Arabic and Telugu, comparable to inflectional plural that may divide into greater, which gives abundance in quantity, and paucal, which yields paucity in number. This implies that languages that have the lexical plural category should fall into the following paradigm: languages with lexical greater and those with lexical paucal.

Keywords:

• lexical plural
• lexical paucal vs. lexical greater
• paucity vs. abundance
• non-count nominals
• Jordanian Arabic

1. Introduction

Inflectional plural often counts. In the generative literature (Abney, 1987; Bernstein, 1991; Ritter, 1991, 1995; Valois, 1991), it is treated as a head of Number Phrase (#P) in the extended morphosyntactic projection of nominals.¹ However, Borer (2005) proposes that plural marking could divide. On this basis, inflectional plural can be counting or dividing. The former merges in #P under the counting head #⁰, whereas the latter merges in Division Phase (DivP) under the dividing head Div⁰, as shown in (1) (see Mathieu, 2012, 2014; Gillon, 2015 where it is reported that the counting plural and the dividing plural may coexist in same language).² The main difference between the counting plural and the dividing plural is semantic. The former gives rise to exclusive interpretation. It excludes the singular, and therefore it yields two or more reading. In contrast, the latter gives rise inclusive interpretation (Borer, 2005; Mathieu, 2014; Martí, 2020, among others). In other words, it yields one or more reading.

Plural marking is not solely inflectional in natural languages (Corbett 2000). From a morphosyntactic perspective, it does not always merge under the inflectional head #⁰ or Div⁰. Plural marking can also be non-inflectional. Wiltschko (2008) and Butler (2012) argue, from a cross-linguistic perspective, that the plural is not invariably a syntactic head, as it can be adjunctive to a full DP or a smaller constituent in some languages. Evidence to the adjunctive nature of the plural in some languages is based on its optionality and inability to change the label of the constituent to which it adjoins. Wiltschko (2008) argues that the plural marker in Upriver Halkomelem, for example, is a root-adjunct; while Butler (2012) shows that the plural in Yucatec Maya is a DP-adjunct. Thus, plural marking can be an adjunct below or above a categorizing head.³

Additionally, non-inflectional plural marking can be a derivational morpheme under n⁰. This means that it merges as a derivational head at the upper edge of the lexical domain, as in Amharic (Kramer, 2012) and Somali (Lecarme, 2002). In the relevant studies, evidence supporting lexical plural is that its formation shows the following idiosyncrasies: it is not that productive (i.e., it exhibits lexical gaps), and its interpretation is unpredictable and non-compositional (Acquaviva, 2008; Alexiadou, 2011; Kramer, 2012; Lecarme, 2002; Lowenstamm, 2008). Thus, lexical plural derives new meanings and shows formal properties of the categorizing head n⁰ merging with uncategorized roots:

In Amharic, for example, näfs-at, which is the plural of näfs “soul”, can be inflectional or lexical, depending on the point of merger of the plural marker (Kramer, 2012). The yielded interpretation when the plural marker merges under #⁰ is the compositional and predictable meaning souls. On the other hand, the yielded interpretation when the plural marker merges under n⁰ is non-compositional and unpredictable, which is insects.

However, the interpretation of lexical plural is not always highly unpredictable in natural languages. The lexical plural that is compatible with non-count nominals in Greek (Alexiadou, 2011; Tsoulas, 2007) and Innu-aimun (Gillon, 2015), for example, does not yield the predictable counting reading, yet it always yields abundance in quantity. The semantic relatedness between the counting plural (abundance in number) and the measuring plural (abundance in amount) explains why the abundance reading of the lexical plural in these two languages is not highly unpredictable. Consider the following example from Greek:

On the other hand, it is reported in one study, to the best of the researcher’s knowledge, that lexical plural yields the opposite meaning (i.e., paucity in amount) in one language, which is Telugu (Smith, 2016). Note that this variation in the interpretation of the lexical plural should be added to the body of evidence supporting the lexical nature of the plural in some languages.

The current study demonstrates that Jordanian Arabic (henceforth JA), like Telugu, has lexical plural that merges under n⁰ with non-count nominals and yields paucity in quantity. To illustrate, the feminine sound plural suffix -a:t directly attaches to the non-count (mass) nominal majj’ water’ in (4) and measures a small amount of water, as shown in the translation.

Given the abundance and paucity reading of lexical plural reported in some natural languages, a paradigm of lexical plural should be proposed: lexical plural branches into lexical greater and lexical paucal under n⁰, as shown in (5). The former denotes abundance in quantity, as in Somali (Lecarme, 2002), Greek (Alexiadou, 2011; Tsoulas, 2007), Innu-aimun (Gillon, 2015), Amharic (Kramer, 2012), while the latter yields paucity in quantity, as in Telugu (Smith, 2016) and JA. This dichotomy is conceivable, as it is comparable to inflectional counting plural, which branches into greater and paucal, among other types of plural under #⁰ (Harbour, 2014), as illustrated in (5). It is noteworthy mentioning that DivP and the dividing plural are inflectional; however, they are disregarded in (5), as the dividing plural is always inclusive and therefore non-branching.

Moreover, the current study pays attention to a language-specific observation. It shows that the marker of the lexical plural in JA, the feminine sound plural suffix -a:t, is intrinsically a paucal marker. Apart from paucity in quantity with non-count nominals, it yields paucity in number with their singulativized counterparts. In (6b), it targets the singulative of the collective nominal ruzz “rice” and counts a small number of grains. Therefore, it is compatible with numerals, as in (6c).

Thus, the intended paucal reading in JA is determined by referring to the position where the feminine sound plural suffix -a:t lands in the nominal spine. It denotes paucity in quantity (amount) when it merges under n⁰ with non-count nominals, whereas it yields paucity in number when it merges under #⁰ with singulativized nominals:

The current study thus argues against proposals that treat the plural as a category residing in a unique morphosyntactic position. It does not support Ritter’s (1991) proposal that plural marking invariably resides in #⁰ and Borer’s (2005) that all true plurals are dividing and reside in Div⁰. It instead argues with studies that treat the plural as a category with various interpretations and morphosyntactic realizations and distributions in the nominal spine (Tsoulas, 2007; Acquaviva, 2008; Wiltschko, 2008, 2012; Alexiadou, 2011; Butler, 2012; Kramer, 2012; Mathieu, 2012, 2013, 2014; Gillon, 2015; Smith, 2016, among others).

The outline of this paper is as follows: Section 2 introduces the number system in Arabic, with particular attention given to the paucal category. Section 3 is to review studies that report the lexical plural in natural languages and to integrate it in the plural morphology of JA. In Section 4, it is proposed that lexical plural, as a category that exists in some languages of the world, branches into lexical greater and lexical paucal. Section 5 concludes the paper.

2. Number in Arabic

This section provides a quick overview of the number system in Standard Arabic and Vernacular Arabic, and then it focuses on the paucal category.

2.1. Arabic number system

The number system in Standard Arabic (SA) and most Arabic vernaculars is rich. It comprises the following categories: the singular (count noun), the non-count (collective and mass), the singulative, the dual, and the plural (Borer & Ouwayda, 2010; Borer, 2005; Mathieu, 2013).⁴ Consider the words from SA in (8) that represent the categories mentioned above, respectively. Note that the difference between the examples in (8) and their vernacular counterparts is mainly phonological. (coll = collective noun, bpl = broken plural).

The main difference between the singular and the singulative is that the former is part of the Arabic lexicon, whereas the latter is morphosyntactically derived. For example, kta:b “a book” enters the morphosyntactic derivation as a singular noun and therefore needs no division. On the other hand, the source of the singulative is a non-count nominal, which is singulativized within the morphosyntactic derivation by the suffixation of the singulative morpheme -at, as shown in (10c) (Borer, 2005; Fassi Fehri, 2018; Mathieu, 2012, 2013; Zabbal, 2002).⁵

The dual has two forms inflected for case in SA: the nominative -a:(n) and the accusative or genitive -aj(n).⁶, while -e:n is the only dual form used in most Arabic vernaculars.⁷ The dual is morphosyntactic; its formation requires the suffixation of the dual morpheme to the singular, as in (8d). The plural category may target morphologically unmarked singular nominals, as in (8e). It may also target singulative nominals, resulting in the plural of the singulative, as in baqar-a:t “some cows”.

From a morphological perspective, the plural in Arabic has two modes: the concatenative and the non-concatenative. The former is shaped by affixing the masculine sound plural -u:n or the feminine sound plural -a:t to a stem linearly, generating the so-called sound plurals:

Phonologically, the feminine and masculine sound plurals are formed by epenthesizing a long vowel or lengthening an existing short vowel in a singular nominal (McCarthy, 2008, p. 303; McCarus, 2008, p. 243). As shown in bold in (9a), pluralization of the masculine singular nominal seems to be carried out by lengthening the vowel u or replacing it with the long vowel u:. As for the feminine singular nominal in (9b), pluralization is performed by lengthening the vowel of the feminine singular marker. On this basis, the sound plural, whether masculine or feminine, is the product of fusing a marker (the nominative marker or the feminine singular marker) with vowel lengthening.

On the other hand, the non-concatenative mode of pluralization in Arabic is formed by mapping a plural template as a morphological unit, which can be viewed as a prosodic pattern on a singular nominal (stem), producing the broken plural. Arabic comprises about 31 distinct broken plural patterns (McCarthy & Prince, 1990). Consider the formation of the following SA broken plurals⁸:

Note that not all the broken plural templates in SA are available in Arabic vernaculars.

The paucal and greater are components in the domain of the broken plural templates (Fassi Fehri, 2018; 2020Ojeda, 1992) in SA. A paucal or greater plural template can be mapped onto one nominal, such as mapping the paucal template ʔaC₁.C₂uC₃ and the greater template C₁u.C₂u:C₃ to the singular ʃahr “month” in (11). Regarding Arabic vernaculars, no previous study reports such a distinction, as far as I know. However, empirical efforts should be paid to determine whether users of Arabic dialects make such a distinction. (pauc = PAUCAL)

Concerning the plurative, it exists in Standard and Vernacular Arabic. As introduced in Fassi Fehri (2018, 2020), it covers individuals yet is analyzed as one unit, as it is morphologically marked with the feminine singular morpheme -at. Therefore, it may trigger feminine singular agreement with a verb. Consider the plural of nadʒdʒa:r “carpenter” in (12).

The rest of this section explores the paucal in Arabic.

2.2. Paucity in Arabic

Paucity is reported in several studies on Arabic number. As mentioned above, inflectional paucal, which yields paucity in number, is a prominent feature of some broken plural templates with a numeral range between 3 and 10 (Ojeda, 1992; Wright, 1933), such as the prosodic plural templates ʔaC₁.C₂uC₃ and ʔaC₁.C₂iC₃-ah that are mapped onto the singular count nominals in (13) to denote paucity in number⁹:

Additionally, the sound plural can yield paucity in number. The first case is the so-called plural of the singulative (Dali, 2020; Mathieu, 2013, 2014), which can only be formed by suffixing the feminine sound plural -a:t:

The feminine sound plural of a count nominal can be paucal if a broken plural template can also pluralize that nominal. Consider the examples in (15) where the sound plurals are paucal, unlike their broken plural counterparts. This observation implies that when they co-exist, the feminine sound plural of count nominals and their broken plural counterparts generate paucity vs. abundance contrast in Arabic.

Regarding the masculine sound plural, there are few exceptional cases where a nominal can be pluralized either by the masculine sound plural or by a broken plural template. If this condition is met, the masculine sound plural denotes paucity:

The previous discussion emphasizes that inflectional paucal denoting paucity in number can be yielded in SA by the use of (1) specified broken plural templates, (2) the plural of the singulative, (3) feminine sound plurals of nominals with broken plural counterparts, or (4) exceptionally masculine sound plurals of nominals with broken plural counterparts. Concerning JA, it can only use the plural of the singulative to yield inflectional paucal (i.e., paucity in number). Although SA excels over JA in generating inflectional paucal forms and yielding paucity in number, JA has a paucal that SA lacks, namely lexical paucal that denotes paucity in quantity.

In the next section, lexical plural as a plural category in natural languages is explored.

3. Lexical plural

This section begins with an overview of previous work on lexical plural. It discusses cases of lexical plural in some languages where it gives rise to one of the following interpretations: abundance and paucity in quantity. Then, it integrates lexical plural denoting paucity in quantity in the number system of JA. At the end, it gives a hint on the mass-count distinction in JA, as the lexical plural in this Arabic variety targets non-count nominals.

3.1. Overview: lexical plural in natural languages

As mentioned earlier, lexical plural (Acquaviva, 2008; Alexiadou, 2011), which targets non-count nominals, such as water and wheat, is reported in some languages. Most previous studies that report this type of plural demonstrate that it yields abundance reading (a lot of), as is the case in Greek (Alexiadou, 2011; Tsoulas, 2007), and Innu-aimun (Gillon, 2015). Consider the following examples from these languages where the plural of mass nominals yields abundance in quantity:

Worthnoting here is that the plural with non-count nominals yields abundance in quantity in Halkomelem Salish, as in (19); however, this plural is regarded as an adjunct in Wiltschko (2008), as it is optional and cannot change the label of the constituent to which it adjoins. The resemblance between this plural and the plural in the examples in (17) and (18) is that they are all lexical, yet the one in Halkomelem Salish is a root-adjunct. Albeit adjunctive, it yields the same interpretation, abundance in quantity. So far, plural marking at the upper edge of the lexical domain or below it (whether it is a categorizing head n or a root-adjunct) gives rise to abundance in quantity.

Smith (2016) also reports that Telugu has lexical plural; however, it yields paucity in quantity. In (20), the plural of the mass noun paa “milk” in Telugu gives rise to a small amount (some of) reading.

The examples in (17), (18) and (20) from different languages and the example in (21) from JA, which yields paucity in quantity, imply that lexical plural can be the product of pluralizing non-count nominals in some natural languages. Still, it does not yield a single interpretation. It instead falls into a paradigm: it yields abundance in quantity in some languages and paucity in quantity in others. The selection of the intended interpretation, whether abundance or paucity, is language-specific and seems arbitrary.

In the rest of this section, it is demonstrated that JA has a lexical plural that denotes paucity in quantity.

3.2. Lexical plural in Jordanian Arabic

Lexical plural in JA is the product of the pluralization of non-count nominals. Non-count nominals are first introduced in this section. As a category, it branches into collective nominals and mass nominals. In Arabic, collective nominals refer to kinds or groups of objects as a whole without any reference to the internal members (see Erwin, 1963; Harrell, 1962; Mathieu, 2013; Talmoudi, 1980). Collective nominals in Arabic refer to fruits, vegetables, grains, animals, insects, and germs. On the other hand, mass nominals refer to substances and materials with no discrete subsets, i.e., tiny atoms, liquids, gases, dairy products and powders, minerals, and other materials. Consider the examples in (22 and 23) where the feminine sound plural of non-count nominals gives rise to paucity in quantity (a small amount reading) in JA. In (22), -a:t attaches to collective nominals, whereas it is combined with mass nominals in (23).

The plural of a mass nominal can only yield paucity in quantity; therefore, it is incompatible with numerals:

This incompatibility implies that the feminine sound plural morpheme merges directly with a mass nominal:

On the other hand, the plural of a collective nominal can denote paucity in number or quantity. The intended interpretation of the plural of a collective nominal is determined with reference to contextual clues. In (26), the plural of the collective gamuħ denotes paucity in number due to the presence of a numeral to its left. The presence of a numeral entails that the collective nominal is first divided (singulativized), and then the plural morpheme merges in the inflectional domain of the nominal to count.

In contrast, the plural of the collective nominal in (27) indicates paucity in quantity. The context indicates that Speaker 2 (the seller) means that the remaining amount of wheat is small. In this context, a numeral cannot appear to the left of the plural of the collective nominal, as it is not a counting one. Instead, it measures a small quantity.

The intended interpretation is yielded once the feminine sound plural suffix merges in a specific syntactic location. If the suffix merges with the singulativized form of the collective nominal, as shown in (28), it is inflectional counting plural (i.e., inflectional paucal).

On the other hand, if this suffix directly merges with the (non-singulativized) collective nominal under n⁰, as in (29), the resulting plural is lexical, and the intended interpretation is paucity in quantity. This analysis implies that the feminine sound plural suffix always yields paucity with non-count nominals, whether they are singulativized or not.

On this basis, JA has lexical plural, like Greek, Innu-aimun, and Telugu. However, concerning the interpretation of lexical plural JA and Telugu stand in opposition with Greek and Innu-aimun. In JA and Telugu, the only possible interpretation of the lexical plural is paucity in quantity, whereas it gives rise to abundance in quantity in Greek and Innu-aimun. Note that notwithstanding its adjunctive nature, the plural in Halkomelem Salish can be classified with that of Greek and Innu-aimun, as it adjoins within the lexical domain and yields abundance reading.

This dichotomy may imply that the semantics of the plural morpheme determines the yielded interpretation in a language. In JA, for example, the meaning of the feminine plural morpheme -a:t is typically paucal, whether it merges in the lexical or the inflectional domain of a nominal. What argues with the selection of the morpheme -a:t to be the marker of paucity per excellence in JA is that it is also the only marker available to pluralize the singulativized forms of non-count nominals, yielding paucity in number. Likewise, it denotes paucity with count nominals that broken templates can pluralize. Thus, the morpheme -a:t yields paucal reading in the spine of non-count and many count nominals.

3.3. Lexical plural and mass-count distinction

This part of the section is intended to answer the following question: does the existence of the lexical plural have any implication for mass-count distinction in JA? The compatibility between plural morphology and non-count nominals (especially mass nominals) is observed in JA; however, this compatibility is not full. Only one component of plural morphology in JA can combine with non-count nominals, namely the morpheme -a:t. Thus, JA, akin to Standard Arabic, distinguishes between count and non-count nominals (especially mass). In other words, both Arabic varieties have a grammatical mass-count distinction. The following paragraphs illustrate the morphosyntax of pluralization in JA.

Singular (count) nouns are freely pluralized. Pluralization is carried out in their inflectional domain. If the plural is dividing (Borer, 2005; Mathieu, 2012), it has inclusive reading. In this case, the plural morpheme merges under Div⁰, but it may move up to #⁰ to perform counting, such as the masculine and feminine sound plural of the count noun muʕallim in (30).

Concerning the plural with exclusive reading, no division is performed, and therefore the plural morpheme merges under #⁰, such as the inflectional templatic paucal and greater in (31a,b):

On the other hand, the non-count category branches into collective and mass. Collective nominals can be inflectionally pluralized exclusively by -a:t once they are divided, as shown in (28) above. In this case, the plural is to count small numbers (i.e., paucity in number). In contrast, no division is performed with mass nominals or undivided collectives, as in (25) and (29). Therefore, this type of pluralization, which is exclusively carried out by -a:t, is not to count. It measures a small quantity.

The previous analysis entails that lexical plural JA is exclusively a function of the feminine sound plural morpheme -a:t. It is the only plural morpheme compatible with non-count nominals in this dialect, yet it cannot count in this case. It can be inferred from these observations that lexical plural does not pose a problem to the proper mass-count distinction in JA.

The following section suggests branching lexical plural into paucal and greater, comparable to inflectional plural under #⁰.

4. Branching lexical plural

Given the variation in the interpretation of lexical plural in natural languages, it is suggested in this section that lexical plural, as a category that exists in some languages of the world, branches into lexical greater and lexical paucal. The former denotes abundance in quantity, whereas the latter gives rise to paucity in quantity. This branching, in turn, feeds the plural category; it can be inflectional/high or lexical/low. Inflectional plural, among other types of plural, embeds the inflectional greater and paucal. Likewise, the lexical plural branches into the lexical greater and paucal. Inflectional plural targets singular or singulative nominals and denotes numbers, i.e., inflectional greater and paucal operate within the inflectional domain of atoms and merge above the categorizing functional projection of a nominal. On the other hand, lexical plural targets non-count nominals (or may be roots, such as the adjoining plural in Halkomelem Salish) and refers to quantities. Precisely, lexical greater and paucal operate in the lexical domain of non-atomic nominals (mass nominals) or not yet atomized nominals (collectives) and merge under n⁰:

All four inflectional and lexical plural types are attested in natural languages: (1) SA has inflectional paucal and greater, (2) JA and Telugu have lexical paucal, (3) Greek and Innu-aimun (and Halkomelem Salish) have lexical greater. A language may have inflectional paucal and greater, yet no evidence is provided for the existence of lexical paucal and greater in one language so far. The lack of such evidence can be attributed to the arbitrariness and idiosyncrasies of the lexical domain. Determining the interpretation of the lexical plural at the lexical level in a particular language, whether lexical greater or lexical paucal, should indicate that language users arbitrarily select only one of the two possible interpretations. Another possible justification is that the plural marker itself intrinsically denotes paucity or abundance, such as the feminine plural suffix -a:t in Arabic, which is by default a paucal marker and can only produce nominals denoting lexical paucal (paucity in quantity).

The previous discussion indicates that the full proposal of the inflectional plural and the lexical plural in the current research is shown in (33), where all the types of the inflectional plural and the lexical plural are observed in natural languages:

The branching lexical plural in (33) implies that further efforts should be made to explore other natural languages containing lexical plural with abundance or paucity interpretation.

5. Conclusion

This paper calls for branching lexical plural into lexical greater as in Greek and Innu-aimun and lexical paucal as in Jordanian Arabic and Telugu, comparable to inflectional greater and paucal. Hence, languages with lexical plural should fall into the following paradigm: languages with lexical greater and those with lexical paucal. The observation that lexical plural in one language can have only one of these two interpretations calls for the complementary distribution of the lexical greater and lexical paucal; each exists in a set of languages. The selection between these two interpretations is language-specific.

This paper has also demonstrated that the feminine sound plural marker -a:t is a marker of paucity per excellence in Arabic. It yields paucity in number if it merges in the inflectional domain of a singular or singulativized nominal. In contrast, it denotes paucity in quantity if it merges under n⁰ in the morphosyntactic projection of a non-count nominal. Another issue raised in this paper is that the existence of lexical plural with non-count nominals does not question the mass-count distinction in JA. This dialect has mass-count distinction, yet non-count nominals can exceptionally be pluralized at the lexical level by the feminine plural marker -a:t to measure a small quantity.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Abdulazeez Jaradat

Abdulazeez Jaradat is an Associate Professor of Linguistics at Applied Science Private University, Amman, Jordan. His main interests are in syntax-phonology interface, morphosyntax and grammaticalization and language change.

Notes

1. Note that NumP does not exist in all languages (Wiltschko, 2008), as in Korean (Kwon & Zribi-Hertz, 2004).

2. To Borer, DivP is a classifier phrase.

3. Note that Wiltschko (2008) and Butler’s (2012) evidence to where a plural can adjoin in the spine of a DP is not reviewed in this paper as it is beyond its scope.

4. Arabic number system, its components and its morphosyntax constitutes a growing research domain in Arabic (cf. Benmamoun et al., 2014; Borer & Ouwayda, 2010; Dali, 2020, 2020; Fassi Fehri, 2018; Jaradat & Jarrah, 2022; Kihm, 2003; Lahrouchi & Lampitelli, 2014; Lahrouchi & Ridouane, 2016; Mathieu, 2013, 2014; Ojeda, 1992; Zabbal’s, 2002).

5. The singulativization of a collective nominal (e.g., fruits and grains) results in an unitization reading, whereas the singulativization of a mass nominal (e.g., minerals) gives rise to partition reading (Fassi Fehri, 2018; Jaradat & Jarrah, 2022). Non-count nouns can also be divided by classifying or measure phrases in Arabic, such as ħabb-at tamr “a grain of dates”.

6. From a historical point of view, the bound dual -a:n and -e:n are the grammaticalized forms of lexical ʔiθna:n, given related phonological clues. ʔiθna:n underwent phonological attrition (Norde, 2012; 2019Traugott, 2002) and were turned into morphologically bound dual morphemes (i.e., ʔiθna:n > -e:n).

7. The dual can be conveyed lexically in Arabic.a. ʔiθna:nʃaajTwotea“two cups of tea”

8. How this mapping operation takes place is debatable. For instance, Hammond (1988) and McCarthy and Prince (199a,b) argue for the melodic transfer accounts whereby targets are categorized words. On the other hand, Kihm (2003), Lahrouchi and Lampitelli (2014) and Lahrouchi and Ridouane (2016), among others, argue for the vocalic insertion accounts whereby the broken plural templates should have access to roots.

9. Broken plural is part of the Arabic non-concatenative inflectional morphology. It is formed by inserting vocalic segments into the consonantal segments of the singular. This operation changes the quality of the vowels of a singular.