Quantity expressions in the Gumer variety of Gurage

Abstract

The aim of this study was to investigate how entities are conceptualized and quantified in the Gumer Variety of Gurage, a South Ethiosemitic language. It specifically uncovered the linguistic means used to express quantity and how the quantity expressions interact with a noun head in an NP, the predicate verb, and grammatical categories, such as gender, number, and definiteness. The topic is one of the marginalized areas as we do not find investigation on Gurage languages and most Ethiopian languages. We used a cross-sectional qualitative methodology and a semi-structured key informants interview method. We found that nouns in the Gumer Gurage are not inflected for a number morphologically; thus, bet “house” can be singular or plural. There are, however, lexically distinct plurals (әrʧ “boy” and deng^ja “boys”; әram “a cow” and әraj “cows”). Numerical quantifiers and definite articles are also used to show plurality. The language distinguishes count and mass nouns semantically but not structurally. The count nouns require measure words in addition to the numeral quantifier as in (sost sin k’awa [three cup coffee “three cups of coffee”]), but not *sost k’awa. Syntactically, a numeral quantifier precedes classifiers, and a head noun follows. A quantifier agrees in number and gender, when the noun is human, with nouns and verbs in a clause.

Keywords:

• Gumer Gurage
• Quantity expression
• Semantics
• Syntax

PUBLIC INTEREST STATEMENT

Quantity expressions in the Gumer Variety of Gurage can be suitable for readers and researchers from Cultural Studies as it shows how conceptualization and expression of quantity vary across cultures and languages. This in turn can contribute to scholars from cultural theory who may use it for comparative purposes in developing and testing their theories. People working on gender can find it relevant as this study considers the interaction between gender, quantifiers, noun heads, and the predicate verb in a clause. Descriptive and applied linguists will find it important as a reference in developing typological linguistic theory and/ or for language teaching purposes, particularly to the target population. Scholars from semantics and pragmatics may find it interesting as the study expresses semantic extensions, and cultural specificity of linguistic expressions in addition to shared cross-cultural conceptualizations.

1. Introduction

1.1. Background

Gurage refers to the South Ethiosemitic language speakers living in the Gurage Zone of the Southern Nations, Nationalities, and Peoples’ Regional State of Ethiopia. The administrative capital of Gurage Zone is Wolkite located 150-km southwest of Addis Ababa. Gurage people are largely farmers, reflected in the measurements used, engaged in breeding animals such as cows, horses, mules, sheep, and goats, and cultivating the Ensete (Ensete ventricosum), barley, maize, peas, and beans. They are also well-known traders.

Gumer is one of the 15 Gurage varieties, namely, Chaha, Ezha, Gyeto, Gura, Inor, Ennar, Endegagn, Muher, Mesqan, Dobbi, Kistane, Wolane, Zay, and Silte (Hetzron, 1972). The degree of intelligibility among the languages varies from mutual intelligibility to unintelligibility (Gutt, 1980; Menuta, 2015). The Gurage varieties, excluding Silte and Zay, are spoken by 1,867,377 million people (Central Statistics Authority CSA, 2007). The Zay is spoken in the Islands of Lake Zway of the Oromiya region. Galila which was spoken in Ambo around Lake Wenchi, and Mesemes disappeared, are other Gurage varieties. Amharic, the language of wider communication in Ethiopia, is spoken as a second language. Many people in Gurage towns also acquire Amharic as their first language. There are also pockets of native Amharic speakers in the Abeshge district who were settled during the Dergue regime (1974–1991). Amharic functions as a language of administration, education, and media in the Gurage Zone. Figure 1 shows the administrative map of the Gurage Zone and the areas Gurage varieties are spoken.

Figure 1. Administrative map of Gurage Zone.

Source: https://www.google.com/search?q=Gurage+language+map

In addition to the Gurage varieties, the Cushitic languages Qabena and Libido are spoken in the Zone. Other Cushitic languages, such as Oromo in the North and East, Halaba in the South, Hadiyya in the South West, and an Omotic language, Yem in the North West (cf. Figure 1) enclose Gurage varieties.

The Gurage varieties are sub-grouped as North Gurage (Kistane and Dobbi), East Gurage (Wolane, Silte, and Zay), and West Gurage (Mesqan, Muher, Ezha, Chaha, Gumer, Inor, Ennar, Gyeto, and Endegagn). The present study focuses on the Gumer Gurage, though discussions are informed by other Gurage varieties. The findings can be transferable to other West Gurage varieties that are genetically closely related (cf. Hetzron, 1972; Menuta, 2015).

1.2. Problem statement

Quantity expression is a marginalized area of investigation in Ethiopia. We found a couple of works, namely, Tesfamicael et al. (2021) and Menuta (2018). The former compares Opo, Aari, and Amharic mathematical concepts, and the latter describes number and numeracy systems in Aari. Aikhenvald (2000, p. 227) also mentions Oromo and Amharic languages regarding the size and weight measurement of masculine and feminine nouns. Though there are several grammatical descriptions of Gurage varieties as for Ezha (Assefa, 2014), Dobbi (Teklemichael, 2002), Ennar (Habte, 2020), Endegagn (Adigeh, 2015), Inor (Abza, 2019), Mesqan (Shafi, 2015), Kistane (Wakjira, 2010), Silte (Oda, 2017), Wolane (Meyer, 2006), and Gumer (Völlmin, 2017), there are no studies specifically addressing quantity expressions in the Gurage languages in general and in the Gumer variety. The present study is required to fill the knowledge gap on quantity expressions in Gumer Gurage. The findings of the study can be used to prepare language and mathematics teaching materials for mother tongue education, which are lacking, in the Gurage Zone. The research is further needed for a comparative typological study of Gurage varieties and other Ethiosemitic languages regarding multiple markings and classifier language groupings.

1.3. Objective and significance of the study

The study aims to uncover how quantity is expressed in the Gumer variety. It specifically attempts to (a) investigate the linguistic means the language uses to express quantity and (b) uncover how quantity expressions interact with a noun head in an NP, the predicate verb, and grammatical categories, such as gender, number, and definiteness.

The findings will be significant for documenting the language, using it in preparing pedagogical grammar, and informing curriculum development in mathematics focusing on indigenous knowledge systems. It can also help for typological purposes, comparing the quantity expressions among Ethiopian and African languages.

2. Methodology

The study follows a cross-sectional qualitative typological description. Linguistic data were gathered from seven key informants in the Arekit town, and Enjefo Kebele of the Gumer district. Three of them were females, and four of them were males. They were native speakers of the Gumer variety and lived in Gumer for more than 10 years. They were selected using purposive sampling in which language experts from the district communication bureau, teachers teaching in Arekit high school, and different gender groups (male and female), and different age groups (with ranges of 26–45 years) were included based on their knowledge of the language and culture. The language variety was selected because it is the least described variety compared to other Gurage varieties. Semi-structured key informant interview was used as a method to elicit linguistic and cultural data. The elicitation largely asked for expressions to individual key informants on the quantity of entity with count, mass, and collective nouns. The elicitation session with the key informants took 30–35 min. The elicited texts and other available written literature were used as data sources. The key informants were informed about the objective of the study and that participation is based on their free will. They were also informed that they can withdraw from the project. The anonymity of the participants was also guaranteed. The collected data was transcribed phonemically, glossed morpheme-by-morpheme with the Leipzig glossing rule, and transliterated into English. Then, each example was discussed regarding the forms and semantic roles they have in relation to quantification.

2.1. Literature review

Quantity expressions are linguistic forms that show how “languages express the fact that reference is being made to a quantitatively delimited amount of the thing mentioned” (Greenberg, 1972, p. 3). Both number words and quantifiers can be used to delimit the quantities of entities. We consider quantity expressions, following Doetjes (2021, p. 54), “to cover expressions such as numerals, many, several, and a lot, but also number markers, measure words, and numeral classifiers”.

Numerals are words, such as one, two, and three that are used for counting, and number words include many, several, little, and few which are used for measuring. Number markers are often affixes added to nouns to express singular, plural, and dual, such as lɨʤ “boy” and lɨʤ-oʧ (boy-PL “boys”) in Amharic. Regarding the measure words and classifiers, authors use different criteria to distinguish them. Her and Chen (2013, pp. 37–46) use semantic and typological criteria. Semantically, a classifier categorizes a noun by describing its core properties but does not add any extra information; a measure word may impart an additional property to the noun categorized. Similarly, Her and Hsieh (2010, p. 527) state that a measure word is “semantically substantive and thus blocks numeral quantification and adjectival modification to the noun; a classifier, in contrast, does not form such a barrier for it is semantically null in the sense that it merely highlights a semantic aspect inherent to the noun and thus contributes no additional meaning.” Typologically, classifiers are used to categorize discrete entities that are countable, but measure words may apply to mass nouns. However, this typology does not seem universal. For instance, Tai (2003, p. 312) argues that “nouns in classifier languages denote materials or substances, non-discrete and unbounded, while in English and other European languages; they denote objects with discrete boundaries.” According to Aikhenvald and Green (1998, p. 430), a classifier “is used to denote a continuum of methods of noun categorization.”

Her and Hsieh (2010) group classifiers into Sortal and Mensural. A Sortal classifier “sub-categorizes objects with a reference to their intrinsic properties, while a mensural classifier measures the quantity” (Her & Hsieh, 2010, p. 528). The authors provide examples such as Yi ben shu (One CL book “one book”)¹ where ben functions as a Sortal classifier and Yi xiang shu (one box book, “one box of books”) where xiang functions as a mensural classifier. These authors state that typologically, “measure words (M) are a mundane part of all natural languages, but Sortal classifiers uniquely set apart a small number of classifier languages, such as Chinese and Japanese” (p. 528).

Typologically, languages use different types of classifiers. Chinese, for example, has three groups of classifiers, namely, the nominal which categorizes human beings, plants, animals, and other objects, the verbal which categorizes actions and events, and the temporal classifier which categorizes time. Aikhenvald (2000) provides grammatical agreement classes, which she calls noun classes, or genders, that categorize nouns based on semantic characteristics, such as animacy, sex, or humanness. Some languages, such as Japanese, have numeral classifiers, morphemes that are affixed to numerals, or quantifiers that categorize shape, animacy, or other inherent properties of nouns. Tariana, a South American language, has a possessed classifier that characterizes a possessed noun and relational classifiers that characterize how “the referent of a possessed noun relates to that of the possessor” (Aikhenvald, 2000, p. 3). Waris, a Papuan language, has verbal classifiers, special morphemes attached to verbs that categorize a noun functioning as a subject or object in terms of “its shape, consistency, and animacy” (Aikhenvald, 2000, p. 3). Classifiers considered rare typologically are locative classifiers, morphemes added to locative adpositions, and deictic classifiers added to deictics and articles.

English has seven classifiers, namely, collective, varietal, measure, arrangement, unit, number set, and fractional classifiers. Her and Chen (2010, p. 3) argued that fractional classifiers such as two-thirds of N, and number set classifiers such as several thousand of N should be considered quantifiers. They have, thus, collective classifiers, species classifiers (for varietal classifiers), standard measures, arrangement classifiers, unit classifiers, container classifiers, and verbal classifiers.

Classifiers interact with a plural system and gender morpho-syntactically (Aikhenvald & Green, 1998). For instance, a language categorized by a numeral classifier will not have an obligatory plural marking on nouns (Doetjes, 2007; Greenberg, 1977). Similarly, gender interacts with classifiers as some languages classify nouns as masculine, feminine, or neuter. Both number and gender may interact with animacy, another core semantic category of nouns. Gumer Gurage data also show definiteness interacts with number and gender, which interacts with humanness.

An empirical literature on quantity expression pertinent to Ethiopia was found in Tesfamicael et al. (2021), in which cultural mathematical concepts of counting, measuring, locating, designing, playing, and explaining (the concepts adapted from Bishop, 1988, 2002), are discussed in Opo, Aari and Amharic languages of Ethiopia. The counting, measuring, and locating are relevant to the present study. They found that the sampled languages used body parts such as fingers, and the whole body of a person as a means of counting (Tesfamicael et al., 2021, p. 107). Measurements such as space, time, and weight are made with objects such as ropes, wood, and utensils. The weight of some items, such as meat, potatoes, and tomatoes are measured with mәdәb “a space the objects occupy”. The amount of entity considered mәdәd for different objects is a rough estimate and may vary in actual weight (Tesfamicael et al., 2021, pp. 106–107).

Menuta (2018) provides a descriptive account of the number and counting systems in Aari, an Omotic language. The finding showed that the number is expressed with pronouns (pp. 26–28), verb agreement affixes (pp. 28–30), plural marking affixes, such as {−na} attached to nouns, and adjectives (pp. 30–31). Counting is largely associated with human body parts, such as fingers and toes and the whole body. For instance, “boonda is ‘one full person’ implying ‘twenty’ since a healthy person- a person without a disability in his fingers and toes, has 10 fingers and 10 toes, a total of 20” (p.32).

3. Presentations of results

Quantity in the Gumer variety of Gurage is expressed with number markers, numerals, and measure words. Some numerals require the addition of classifiers to quantify the nouns being measured.

3.1. Quantity expression with grammatical number

Nouns in Gumer Gurage can be singular or plural, but neither is marked with a bound morpheme. Some Gurage varieties, however, have grammatical affixes (baj-oʧ [boy-PL “boys” in Kistane, sab-ʧә person-PL “people” in Silte]. Gumer variety expresses plural with lexical items, numerals used as quantifiers, pronominal suffixes functioning as definite markers, and verb agreement affixes in clauses indexing the subject and/or object noun.

3.1.1. Plural expressed with lexical items

In Gumer Gurage, there are very few sets of nouns that show plural with lexical items. Such nouns categorize only [+animate] nouns, specifically human beings, and domestic animals, as shown in (1). We do not find such lexical plurals with plants and inanimate entities.

The plural formation in 1a, ərʧ, and deng^ja, is suppletive. In gərəd and gɨred, 1b, there are vocalic changes, but the consonant roots are the same; əram and əraj in 1c differ in their last consonants/m/and/j/; in mɨʃt and ɨʃta, the/m/of the singular form is deleted and non-existing vowel/-a/is added at the end of the plural form. Such plurals are attested in most Gurage varieties, such as Kistane (Wakjira, 2010), and Dobbi (Teklemichael, 2002). The lexical plurals can be used with or without² numerals to express quantity as in (2).

The example in 2a indicates the speaker has two or more male children, but the subject noun, Tademu in 2b, has specifically three male children. Thus, the lexical plurals show that the noun refers to more than one entity. The addition of a numeral quantifier (2b) limits the plural to a specific number of referents, “three” individuals. A lexical item, such as gɨred “girls”, specifies the noun as feminine, plural, and human, but the numeral modifier sost “three” in sost gɨred “three girls”, limits the number of girls to three, otherwise maybe two, four, etc.

3.1.2. Plural expressed with pronominal suffixes

In some Gurage varieties, definiteness is expressed with grammatical affix as in Kistane and Mesqan with {−i}, gәrәd “a girl” and gәrәd-i “the girl”; Muher and Ezha with {−we} as in gәrәd “a girl” and gәrәd-we “the girl”. In such language varieties, definiteness does not interact with number and gender. Gumer variety on the other hand expresses definiteness with pronominal affixes associated with number and gender as in (3).

The definite marking pronominal suffixes {−hut} and {-hɨno} specify animate and inanimate nouns for singular and plural, respectively. The definite marking pronominal suffixes {-h^jɨt} and {-hɨnәma} specify only [+human] singular and plural feminine nouns, respectively. Inanimate and [-human] nouns are not specified for feminine gender, as shown in (4):

In 4a, taj- “sheep” is an indefinite common noun, but the addition of {-hut} makes it definite, masculine, and third-person singular. In 4b, {-hɨno} suffixed to t’aj- “sheep” makes the noun definite, third-person masculine, and plural. In 4c and 4d, the situation is different; gәrәd “girl” and gɨred “girls” are already specified for feminine singular and feminine plural, respectively, lexically. The pronominal suffixes {-hjɨt} and {-hɨnәma} specify the definiteness agreeing in number and gender to the specified noun and the verb of the clause. The {-әʧ-} in (4c) and {-әma-} in (4d) are agreement affixes for 3fsg and 3fpl, respectively. Note that “sheep” is considered masculine as the definite markers in 4a and 4b, and the verb agreement affix {−jo} in 4b are masculine markers. The language, however, distinguishes the sex difference between taj “sheep” and onә “ram”. All [−HUMAN] nouns are masculine grammatically, despite their animacy as examples 4e and 4f behave the same as 4a and 4b although bet “house” is [−ANIMATE] and [−HUMAN} but taj “sheep” is [+ANIMATE] but [−HUMAN].

3.1.3. Plural expressed with verb agreement

Verbs in Gumer agree with the subject and object nouns (cf. 5f). However, numeral quantifiers do not agree with the noun they modify. Consider the noun bet “house” in (5):

In 5a) the subject noun Mahmud, the object noun bet “house”, the quantifier at “one”, and the verb sɨjәm “he bought” are singular; hence, there is an agreement in number. In 5b, the subject and object nouns are singular, though the object noun quantifier h^wet “two” is plural. The verb in 5b is singular agreeing with the subject noun but not with the quantifier. Note that/r/in the verb sir is a sonorant alternation common in Gurage languages (cf. Menuta & Meyer, 2015). In 5c, the subject NP is plural but the object NP, bet “house” is singular though it is quantified by h^wet “two”. The verb sɨrabom “bought (3mpl)” agrees with the subject NP in number. The fact that the numeral quantifiers do not agree with the number of the quantified object NP is well demonstrated in 5c and 5d where bet “house” remains the same though the numeral quantifiers vary. The object NP modifier h^wet-ʧɨm (two-ncl “both”) in 5e, however, makes the verb agreement plural; hence, we have plural object NP agreement suffix {-no-} attached to the verb. We assume h^wet-ʧɨm “two of them” or “both” differs from the numeral quantifier h^wet “two”. The numeral classifier {-ʧɨm} specifies the numeral, and agreement with the verb is made with the whole object NP (quantifier +N). The numeral classifier {-ʧɨm} can be used with all numbers from 2 to 8 as in sost-ʧɨm “all the three NP”, arbә-ʧɨm “all the four NP”, etc. Numerals above 9 require pronominal suffix {-mәhuna} as in zәt’әmәhuna “nine of them”. Note that *zәt’ә-ʧɨm is not acceptable.

In 5f, the subject of the sentence is singular, and it is co-referenced with {−j}, ‘third person singular masculine; the object deng^ja “boys” is co-referenced with {−bo} “third plural masculine object”. There is no formal difference with the numeral quantifier h^wet whether the object noun quantified has the same form for singular or plural as in bet “house” or a lexically distinguished plural form as in (5f).

Mass nouns such as ɨha “water” t’afi “teff” cannot be counted with one, two, etc. They, however, can be measured with mensural classifiers such as in at wəʃərɨha (one pot water “a pot of water”) and sost k’unna t’afi (three basket teff ‘three baskets of teff). Compare the examples in:

As the ill-formed structures in 6a and 6c and the grammatical structures in 6b and 6d show, the nouns ɨha “water” and t’afi “teff” are not counted in the Gurage culture so must be counted with a container classifier. The details on numeral quantifiers, and the different types of classifiers used are discussed in 2.2.

3.2. Numeral quantifiers and classifiers

The most used classifiers in the Gumer variety include collective classifiers, arrangement classifiers, unit classifiers, container classifiers, and measure classifiers. In some cases, there is no clear-cut boundary between unit classifiers and container classifiers because some containers are used as units of measure.

3.2.1. Collective classifiers

The collective classifier also called a group measure is used to quantify collective nouns such as people, family, etc. as in crowd in the crowd of people, and flock in the flock of birds. According to Zhang (2011, p. 4), a collective classifier is a ‘Unit for counting groups of non-mass elements” such as “shi dui luobo ([ten CLC carrot] ‘ten piles of carrots’)” which in our case considered arrangement classifiers (2.2.2). There are a few collective classifiers in Gurage. A typical example is given in (7):

Classifiers that show pairs are also considered collective classifiers. There are two classifiers showing pairs in Gumer Gurage: t’ɨmd “tame” and ləmʧa “twin” as in (8).

In 8a, dən “single” is contrasted with t’ɨmd “pair” (8b). Semantically, dәn can be used for a part of a load as in a sack of barely laden on one side of horseback as opposed to the other sack of barely in the opposite belly of the horse; but in 8a, it refers to an ox from a tame of oxen. In a general sense, it refers to a “part of a pair”. The collective classifier ləmʧa is used for all animates including humans and animals, such as ləmʧa dәrma (twin pony “twin ponies”). There is an agreement between the collective nouns and the verb. In 8a and 8d, the numeral quantifier is singular “one”. In 8a, we have classifier dәn but not in 8d, yet both mark a single entity. In 8b and 8c, the verb agrees with the singular subject NPs and the plural object NPs (classifier +N). Note that t’ɨmd bora is a PL-SG sequence yielding Plural NP, but ləmʧa gɨred is a PL-PL sequence with PL NP.

3.2.2. Arrangement classifiers

Arrangement classifiers are used to show the constellation arrangements (Xiao & McEenery, 2010) of entities in space. The entities shown can be arrangements of the perceived shapes of the objects described. Some of the common arrangement classifiers in Gumer Gurage are mara “line”, narra “pile”, t’or “load on a head”, t’ɨm^wjət “roll for wusa”, ʤəmb “piled and rolled entity”, and sərat literally “sale” actually “pile of cabbage”. In (9) are sentences with arrangement classifiers:³

In (9a) mara classifies things arranged in horizontal lines, (9b) narra classifies things piled vertically, (9c) t’ɨm^wjət classifies rolls of wusa, (9d) sərat classifies a certain amount of cabbage that are picked, arranged, and tied piling. It is important to mention that ambɨr “cabbage” with the meaning “make one survive” has a strong culture in Gumer, and has diversity with varieties: f^wәt’әra, finʃә, sim^warjә, and gәkәbәtjә. T’or in (9e) classifies piles of entities such as grass, barely, wood, etc. which can be carried by a person on a head at a time.

3.2.3. Unit classifiers

Unit classifiers also called unit words (Tang & Her, 2019, p. 307) are linguistic expressions used as a unit, and they are a natural part of the entity measured. They often classify nouns based on the shape of the entity. A few examples of unit classifiers in Gumer Gurage are given in (10):

In 10a, ʤɨm^wa classifies nouns with round shapes, such as potatoes, onions, garlic, a six-row barley head, etc. In 10b, h^wərəra classifies cereals that have a cover containing the seeds such as peas and beans. In 10c and 10d, both əʧ’ə and k’ut’a semantically refer to “stick”, but the former classifies only a bean’s stick and the latter classifies only a cabbage’s stick. In 10e) əʤ “hand” is semantically extended to the branch of plants; so, in Gumer plants have hands, ‘branches. Thus, əʤ classifies all plants that have branches. Note that əʤ referring to a human body is also used as a measurement referring to a “portion” and as a container classifier.

Some more examples of unit classifiers include ank’wə “ring” as in sost ank’^wә ʃənk^wəra (three ring sugarcane “three rings of sugarcane”), k’jɨnk’ja “pod” as in amɨst k’^jɨnk’^ja wəkaʃə (five pod cardamom “five pods of cardamom”) and k’ɨt’ər “leaf” as in sost k’ɨt’ər ambɨr (three leaf cabbage “three leaves of cabbage”). The word k’ɨt’ər “leaf” is also used as a measure phrase as is asɨr k’ɨt’ər wərək’ət (ten leaf paper “ten leaves of paper”).

Linguistic expressions that are used to measure parts or fractions of the entity measured are also considered as unit classifiers; a few examples of such expressions are given in (11):

Fɨnt, 11a, is used to measure half of the flat or circular solid entities. We can say for example fɨnt wusa (half Wusa “a half of circular and flatbread of Wusa”) and fɨnt dɨniʧa (half potatoes “half of a head of potatoes”). K’ɨrt’ is used to measure flat entities such as wusa a “bread made of Ensete product” and plots of land. While K’ɨrt’ wusa measures one-fourth of any flat-size bread K’ɨrt’ afәr (one-fourth land “one-fourth of a plot”) specifically refers to seven sticks of a plot of land where one stick is about seven arms. The classifiers f^wɨrɨm (11c) and but’ɨr (11d) specifically classify flat-baked bread wusa with unequal proportions and f^wɨrɨm (11c) referring to a “smaller piece” and but’ɨr (11d) to a “bigger piece”.

3.2.4. Container classifiers

A container classifier is a linguistic expression used as containers of different sorts and is used to measure an entity that is designated by a count or a mass noun. Most of the container classifiers are much like what has been categorized as measure words. There are several container classifiers in Gumer Gurage. We grouped them based on the shape of the entity (flat, round, etc.) measured.

3.2.4.1. Body parts as a container classifier

Body part as a container classifier is used to measure food items, drinks, and other objects. It includes ʤef “foot”, zɨsɨr “span”, hɨnə “arm”, en “eye”, dən “belly”, gunər “head”, əgɨr “leg”, k’^jɨn “buttock”, t’ɨb^waʧ’ə “fist”, k’əjə “a handful”, ɨnf^wɨjə “two handful”, əʤ “hand”, ɨnk’^jǝf^wǝ “an armful”, amf^w “mouth”, and finʧ’ə “chicks”. We provided some example sentences with a body as a container classifier in (12):

Some of the body parts are considered container classifiers in the sense that the measurement is made based on the space the entity measured occupies the body parts, such as the distance which is equal to someone’s feet (12a), span (12b), and arm (12c). The body part ʤef “foot” in (12a) classifies a house that is circular and is measured with the foot from one corner to the center of a house where a central supporting pole “әʧba” is erected. The whole distance of a house is, therefore, twice the measurement with feet. Ten feet of a house is twenty feet from one corner to the other ends. The classifier zɨsɨr “span” measures flat entities, such as carpet, rope, table surface, etc. zɨsɨr “span” in 12b measures the size of wədərə “rope” which in turn can be measured with a span containing a butter, which is rolled, piled in a circle, and wrapped with Enwa “a dry stem of Ensete plant”. The classifier hɨnə “arm” is the most common measure of length used to measure flat entities including ropes, lumber, carpets, etc. Two arms are 1 m in modern measurement. Carpets made of Ensete products in Gurage are sold based on the size measured with arms and the quality of weaving. The word en “eye” (12d) is used to measure distance. The phrase en әrәg^wәt “eye throw” is the maximum distance one⁴ can visualize horizontally from the standing deictic center. The eye as a measure of distance is also found in a riddle as in (12o)⁵

The term dən “belly” (12e) is used to classify nouns in two contexts. It can be used with the meaning of “half” of a load as in dən ʧ’ərət (belly load “half of a load”) or “a pair” as in dən bora (belly ox “a single ox”); hence, requires another ox for taming. The general concept of a belly as a quantity measure is “part”. The body part gunnər refers to “head” and “hair”, so the difference is understood contextually. For instance, gunnәr gɨnd (head big “big-headed) or gunnәr m^wɨrt” (hair shaved “bald”). Thus, gunnәr used as a classifier in (12f) refers to “head”, not to “hair”. The word əgɨr “step” classifies flat space, such as a walking distance as in directing someone in asramɨst әgɨr wәr (fifteen leg (step) go “travel fifteen steps”). The classifier k’^jɨn “buttock” (12 g) is used in counting the corm of Ensete. So, sost k’jɨn wәta (“three buttock corm ‘three roots of corms’) refers to three full corms of Ensete which are uprooted from the ground. For potatoes that are under the soil, k”^jɨnk’^jә “pod” instead of k’^jɨn “buttock” is used. Size and shape are important factors for such differences. The corm of Ensete is big like a buttock and is spherical, whereas potatoes are comparatively smaller and typically round though other shapes are available; hence, “pods.” More research on the conceptualization of these two is still required.

The classifiers t’ɨb^waʧ “ə ‘fist’, k’əjə ‘a handful’, and ɨnf^wɨjə ‘two-handful’ are used to measure cereals. They are often used to measure roasted cereals provided to people. The word əʤ ‘hand’ is used as a container classifier in the context of handling such things as barley and grass between the thumb and the index finger. On the other hand, ɨnk”^jǝf^wǝ “armful” is used as a measure of entity, such as sticks of a bean, mowed barley with its straws, etc., that can be handled with a hand bent towards the shoulder.

The classifiers amf^w “mouth” (12n) and finʧ’ə “chicks” (12o) are used as containers to measure the amount of food and drink, respectively. Note that the verb drink differs for coffee and other liquids. For coffee, tәk’aw “you (M) drink, or tәk”^jaw “you (F) drink”, and for other liquids sɨt’e “drink” as in eb sɨt’e (milk drink “drink milk”) is used. The verbs tәk’^jaw and sɨt’e classify hot and cold drinks, respectively. The verb fiʧ’ “sip” specifies a small amount of drink’.

In Gumer Gurage, objects measured with body parts, in turn, can become units of measure. The words ʒɨr “seven arms of wooden stick” and wədərə “a rope” are cases in point. The classifier ʒɨr is used to measure plots of land. The smallest measure of plots of land is ʒɨr afər (stick land “a stick or seven arms of land”); wədərə “rope” is also used to measure land. At wədərə afər (one rope land “a rope of land”) is equal to four ʒɨr “four sticks” or “28 arms long”. Wədərə “rope” is also used to measure k’ɨb “butter” as in wədərə k’ɨb (rope butter “a rope of butter”) which is about a kilogram. In this latter context, a rope has one span length unlike a rope used to measure land which has 28 arms long.

3.2.4.2. Clay-made container classifiers

Clay-made containers are the most common classifiers for measuring liquids such as butter, milk, coffee, homemade beer, and water. In Western culture, butter is considered solid but in Gumer Gurage, it is liquefied by melting and then used in the same way oil is used. The examples in (13) show liquid measuring container classifiers:

The containers that classify the liquids in 13a to 13f are used with singular noun heads and singular verbs in a clause. The classifiers in most cases are restricted to a noun modified. For instance, wəʃər “pot” is not used with k’ɨb “butter” so * wəʃər k’ɨb is not common. Similarly, gan “pot” is not used with eb “milk” or k’ɨb “butter”. In 13e and 13f, want’ɨjə and ʤәbәn refer to “kettle”, but the former is used with butter and the latter with coffee restrictively.

3.2.4.3. Bamboo-made container classifiers

Different size bamboo-made baskets are used as container classifiers. The baskets used to measure objects from the smallest size to the biggest, respectively, are mant’ret “filtering basket”, t’ək’ujə “small basket used to measure grain and potatoes, k’unna” basket containing about 17 kilograms’, k’ərʧ’at“a big basket” and ʃat “granary”. The examples in (14) are their use in sentences:

The container mant’ret is specifically used to measure cottage cheese called k^jәsa. The containers t’әk’ujә, k’una, and k’әrʧ’at can be used to measure cereals and potatoes. The container ʃat is specifically used to measure cereals. The two basket varieties, t’ək’ujə, and k’ərʧ’at, are smeared with animal feces, dried, and then used to remove away urine and dung from animals’ cages. They also assume different names; the smeared t’ək’ujə is called jɨbanguna and used to classify urine, and smeared k’ərʧ’at becomes gurə, and is used to classify dung as shown in (15):

3.2.4.4. Other containers

The leaf of Ensete is warmed with heat and cut in a circular shape, and then used to eat with minced meat called kitf^wə; the shaped leaf is called t’ɨləs, and it functions as a classifier as in sost t’ɨləskɨtf^wə (three CNC kɨtf^wə “three Tiles of minced meat”).

The word g^wəʤə “hole” is used as a container classifier to measure objects placed in it, such as sost g^wəʤə wusa (three hole wusa “three holes of Wusa⁶”), and sost g^wəʤə dɨniʧa (three hole potatoes “three holes of potatoes”). In the former case, g^wəʤə is an actual hole dug to store fermented wusa- a scraped mix of Ensete products, but in the latter g^wəʤə does not refer to the actual hole but to sets of potatoes found under the soil of a single planted potatoes plant which can be dug out for consumption. In any case, the entity is contained inside the soil. Wusa is also measured with jɨtəf^wəp^wə a “small mat” with which raw wusa is shaped into a flat and circled for subsequent baking. As one can bake several of such shaped wusa, jɨtəf^wəp^wə is used as a measure word as in h^wet jɨtəf^wəp^wə wusa (two mat wusa “two mats of wusa”).

Grains may be measured with containers, such as ʤunja “sack” as in ʤunja əhɨr (sack barley “a sack of barley”). Barley which is not threshed is also measured with k’^wat’ərə “tie” as in sost k’^wat’ərə əhɨr (three ties barley “three ties of barley”) where the rope that tied the barley with its straws is considered a container.

3.2.4.5. International measure classifiers

In Gumer Gurage, some of the international measure classifiers, such as milligrams, and centimeters are rare though units such as litter and meter are now becoming common. The two most used ones are shown in (16):

It is important to note that 50 kg and 100 kg of entities can be measured with sacks called k’umt’a and kuntal, respectively. What is more, a jug is used to measure liquids instead of a liter. This implies that both liter and kilogram are only recent substitutions to the indigenous measure classifiers.

3.3. Quantity expressed with postposition

Postpositions are used as quantity expressions of rough estimates and ranges of entities measured with numeral quantifiers as in (17):

In 17a, the speaker promises to serve people whose range may be between one and ten; in 17b, it is estimated that Kasa shall come until three o’clock, but maybe earlier or with a slight delay, and in 17c, Kasa waited for Almaz for about three hours (the beginning and the ending of the time is not specified).

Rough estimates may also be made with numeral quantifiers which are modified by other degree adjectives, such as bәanəsə “at least” and bəbəssa “at most” as in the example in (18):

3.4. Adjectives as quantity expression and classifiers

Quantity may also be expressed with evaluative quantifiers, such as all, few, some, and many as in (19):

The evaluative quantifiers ɨnɨm “all”, kəri “a few”, and bɨzə “many” agree with a singular verb as can be seen from the examples in 19a, b, and c. In this sense, the evaluative quantifiers function as distributive pronouns in English such as each which agree with a singular verb. It is, however, possible to make the verb plural if the focus is on the number not on the act as in (20):

The quantity expression used in Gumer Gurage is summarized in (21). We represented degree marking expressions such as at least, and at most with ∞DQ, numeral quantifiers with ^β NQ, classifiers with ^γ CL, and a noun head or adverb with ^δ NH. Hence, we have the following maximal arrangements:

In (21a) degree quantifier bә-anəsə “at least” comes before the numeral h^wet “two” which comes before the classifier əʤ “hand”, which all contributes to the amount of the head noun əhɨr “barley”. In (21b), we have only numeral quantifier sost “three” quantifying the noun mәt’af “book”; and in (21c), we have numeral quantifier at “one”, a classifier mant’ret “basket” and the head noun k’^jəsa “cheese”. Depending on whether the head noun is mass or count, a classifier can be obligatory (21a and 21c) or optional (21b).

4. Discussion and conclusion

4.1. Discussion

In this study, we investigate the linguistic means the Gumer variety of Gurage uses to express quantity, and how the quantity expressions interact with a noun head in an NP, the predicate verb, and grammatical categories, such as gender, number, and definiteness. We found that there is no grammatical number agreement between a classifier and the noun specified whether the noun is semantically a count or a mass. Thus, the distinction between count and mass nouns seems less significant in the language in terms of a number. For instance, sost angədo waga (three C cattle “three cages of cattle”), sost sin k’awa (three cup coffee “three cups of coffee”), and sost k’ut’a ambɨr (three c cabbage’ three sticks of cabbage) all lack number marker, hence, are treated as a mass noun. This conforms with Greenberg (1972, p. 17) who states “numeral classifier languages generally do not have a compulsory expression of nominal plurality, but at most facultative expression,’’ and with Tai (2003, p. 312) who claims that there is no count-mass distinction in classifier languages. Gumer Gurage is not a plural marking language; it has no overt morphological plural marker on a noun, characteristics of languages described in (Haspelmath, 2013); thus, must use either sortal or mensural classifiers for quantity expression. This conforms to Tang and Her (2019) who claim that sortal classifiers and morphological plurals cannot co-occur and that if a language has both sortal classifiers and morphological plural, the two have to be in complementary distribution. In some languages such as English, measure terms can have plural forms as in three bottles of Wine (Tang & Her, 2019, p. 320), but in Gumer Gurage the measure terms do not have plural markers as in sost t’ərmwɨz wəjn (three bottle wine “three bottles of Wine”) which is structurally like at t’ərmwɨz wəjn (one bottle wine “a bottle of Wine”).

Nouns functioning as subject or object agree in number when the number is marked with lexical plurals or pronominal affixes functioning as a definite marker, and gender with a predicate verb; however, as the default gender is masculine, only [+ANIMATE] nouns have an agreement for feminine gender.

Regarding the structural arrangement of quantifiers and head NP, the quantifiers of Gumer Gurage come first then the classifiers, and finally the head NP. There are, however, exceptions to this default structure. This conforms to the Her (2012, p. 1211) claim that languages may have diverse structures though there is one default form that captures most of a structure as is the case in Chinese.

4.2. Conclusion

Gumer Gurage uses several means to express quantity, and the study showed the following as the major ways:

• The language marginally uses lexical plural to quantify nouns (example-1)

• Pronominal suffixes {-hɨno} ‘them (M) and {-hɨnәma} “them (F),” which are also associated with definiteness, and are used to show plural (example-3).

• Verb agreement pronominal suffixes that index the subject and object nouns express quantity (example 5).

• Numeral quantifiers and classifiers are the main means to express quantity in Gumer Gurage. Numeral quantifiers are used to quantify count nouns (sost deng^ja [three boys “three boys”]), and together with classifiers to quantify collective and non-count nouns.

• Gumer Gurage makes use of several types of classifiers for quantifying, such as collective classifiers (2.2.1), arrangement classifiers (2.2.2), unit classifiers (2.2.3), container classifiers that include body parts (2.2.4), clay made materials (2.2.4.2), bamboo made materials (2.2.4.3), and international measure classifiers (2.2.4.5).

• Rough estimates are expressed with postpositions (example 17) and degree adjectives (examples 18 and 19).

• Regarding the interaction of noun heads and their quantifiers vis-à-vis gender and definiteness, the quantifiers are not inflected for gender (sost dengja [three boys “three boys”; sost gɨred [three girls “three girls”]). Definiteness expressed with pronominal suffixes, however, is sensitive to both gender and definiteness (gәrәd-hjɨt [girl-3SGF. DEF “the girl” and gɨred-hɨnәma [girls-3PLF.DEF “the girls”]).

• As to the arrangements of quantifiers against a noun head, we found the maximum possible order of degree quantifier (DQ), Numeral quantifier (NQ), Classifier (CL), and noun head (NH).

• Finally, we can group Gumer Gurage typologically to a classifier language rather than a plural marking language as it has no overt morphological plural marker on a noun. This can also be true to most central West Gurage varieties, such as Chaha, Inor, and Ezha, but not to North Gurage languages such as Kistane which have overt plural marking morphemes.

Based on the findings, we suggest curriculum planners and teaching materials developers for Gurage language varieties consider the differences between plural marking language varieties such as the Kistane and classifier language varieties such as the Gumer. As the present study focused on the description of a single Gurage variety, Gumer, a comparative study of all the Gurage language varieties can help for typological characterization of plural marking and classifier languages of Ethiosemitic languages.

Acknowledgments

We would like to thank our language informants for their precious time and friendly support during fieldwork. We also extend our gratitude to Hawassa University for financial support through the Linguistic Capacity Building: Tools for the Inclusive Development of Ethiopia project.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Fekede Menuta

Fekede Menuta published several related works on Gurage languages. Menuta (2021), Ensete in Gurage: Nomenclature, use and meaning extension deals with how the Ensete culture contributed to lexical development and metaphorical extensions. Menuta (2022), Verbal Extension and Valence in Gumer Gurage describes how the extension of affixes in the derivation of verbs interact with an argument of the verb and the number of valence. Menuta (2019), Language Use in Multilingual Ethiopia explores language policy and practice among languages, including Gurage, in the Southern Nations, Nationalities, and Peoples Regional State. He provides language used in education, media, administration, and the challenges faced in implementing the multilingual language policy. Menuta (2018), Number and Numeracy in Aari describes how people count and quantify entities. Menuta (2016), Time in the Gurage Variety of Gumer describes how Gumer Gurage speakers conceptualize time and how they express the concept with verbs, adverbs, and even nouns.

Notes

1. ACC=accusative, ARC= arrangement classifier, BEN=benefactive, CAUS=Causative, CL=classifier, CLC= collective classifier, CNC= Container Classifier, CONV= Converb, COP= Copula, DEF= Definite, DQ= Degree Quantifier, F=feminine, GARD= Genitive, Accusative, Relativizer, Dative, IMPF= Imperfective, INDF= Indefinite Future, INST= Instrument, LOC= Locative, M= Masculine, MAL= Malefactive, NCL= Numeral Classifier, NH= Noun Head, NOM= Nominative, NP= Noun Phrase, NQ= Numeral Quantifier, O= Object, PERF= Perfective, PL= Plural, POSS= Possessive, PST= Past, SG= Singular, UC= Unit Classifier.

2. The morpheme {−m} is used to mark past tense, converb, main verb, and focus (cf. Menuta, 2022, p. 95).

3. Sim^wa is the second stage of Ensete transplantation (cf. Menuta, 2021 doi 10.1075/impact.48).

4. k’әjә refers to a single handful () and ɨnf^wɨjəto a “two-handful” ().

5. The morpheme {jә-} in Gumer Gurage functions as an accusative, genitive, dative, and relative clause marker (Menuta, 2019). The different uses of the form are understood contextually based on the syntactic position of a noun and a verb to which the morpheme is attached and other linguistic and pragmatic contexts. We, therefore, used GARD to represent any of the functions of the morpheme marks.

6. Wusa is the product of Ensete obtained by scrapping the stems of the plant; it is stored in a hole for fermentation and subsequent use. The same term is used for the baked bread of the product; hence, the raw and baked form are distinguished by context.