Gàidhlig / English
LEACAN BLOG 2: Gaelic compound words and hyphenation (2)

LEACAN BLOG 2: Gaelic compound words and hyphenation (2)

Posted by Robby on 9th October, 2023
This is the second blog associated with the LEACAN project, funded by Bòrd na Gàidhlig. LEACAN is concerned with developments in the corpus of Scottish Gaelic and how these developments are coordinated and agreed. It is an inter-university project involving the Universities of Glasgow and Edinburgh and Sabhal Mòr Ostaig / UHI, originally under the auspices of the now dissolved Soillse research network for Gaelic language and culture. The purpose of our blogs is to share some of our research findings and to gather comments and feedback on trends, developments and recommendations arising from the work. This our second blog is concerned with certain types of compounds and the related question of hyphenation in such compounds.

Right-branching compounds
Syntactic structures are described as ‘right-branching’ if their head or generic element is followed by its modifiers, i.e. the modifiers are to the right of the head, which is written in bold in the following examples:
  • (compounds) tìr-mòr ‘mainland’, taigh-òsta ‘hotel’, neach-lagha ‘lawyer’
  • (phrases) duine mòr ‘a big man’, taigh cloiche ‘a stone house’, glainne uisge ‘a glass of water’.
Right-branching compound words are normally hyphenated in Scottish Gaelic, although this practice has not always been regular: older texts contain numerous examples of compounds that are written as two separate lexemes (words) without a hyphen.

Additionally, unlike left-branching expressions, right-branching word combinations are often ambiguous in that it can be difficult to determine whether they are syntactic phrases or compound lexemes. One reason for this is phrase structure: most syntactic phrases are right-branching in modern Gaelic.

Right-branching compounding is productive in Gaelic and numerous modern terms have been coined in this way recently to meet the needs of the modern world. These, however, are still not well-established in the language and, hence, tend to be perceived as syntactic phrases rather than compounds, and are thus often spelt as separate words without hyphens.

Due to all of this, it can be difficult to decide whether such an expression should be interpreted as a compound and hyphenated or not.

There are a number of criteria that can help to determine whether a right-branching expression ‘qualifies’ as a compound word or not:
  • the degree of phonetic reduction in the head element
  • head or modifier omission in coordination
  • position of determiners and adjectives
  • pluralisation of the modifier
  • use of the genitive case
  • if its meaning is different from the meanings of its constituent parts
  • hypernymy (explained below)
  • use of preposed intensifiers
  • gender: in at least one compound, the gender does not reflect the current gender of the main noun element, e.g. masculine tìr-mòr ‘mainland’ (cf. feminine tìr).
The head element of compounds tends to be reduced to a certain degree, albeit not consistently. These reductions are indicated in the phonetic transcriptions in square brackets which precede the following examples:
  • [təˈb̥eɡ̊] taigh-beag ‘toilet’ vs. [ˈtɤjˈb̥eɡ̊] taigh beag ‘small house’
  • [rˠad̥ˈmoːr] rathad-mòr ‘highway’ vs. [ˈrˠa-əd̥ˈmoːr] rathad mòr ‘big road’.
When several plural syntactic phrases contain an identical element and occur in a coordinated structure, that element can be omitted from all phrases but one. This appears to be limited in compounds:
  • craobhan giuthais is (craobhan) daraich ‘pine (trees) and oak trees’
  • taighean-seinnse ’s (taighean-)bìdh ‘pubs and restaurants’
  • madaidhean-ruadha ’s madaidhean-allaidh ‘foxes and wolves’
  • *madaidhean-ruadha ’s allaidh (where * indicates an unacceptable formation).
Unlike syntactic phrases, compounds cannot generally be divided by adjectives:
  • craobh mhòr ghiuthais / craobh ghiuthais mhòr ‘a big pine tree’
  • taigh-òsta mòr ‘a big hotel’
  • *taigh mòr òsta.
Determiners (the definite article and possessives) normally precede both compounds and phrases that they modify. However, when the modifying element refers to a specific entity it (rather than the whole expression) is preceded by the determiner. Such expressions are interpreted as syntactic phrases:
  • (compounds) an taigh-seinnse ‘the pub’; m’ eachdraidh-beatha ‘my biography’
  • (phrases) am plana leasachaidh ‘the development plan’
  • (phrases) bean an taighe ‘the landlady’ (lit. ‘[the] woman of the house’); bràthair mo mhàthar ‘my maternal uncle’ (lit. ‘my mother’s brother’); comhairle a’ bhaile ‘the town council’; mullach an taighe ‘the roof’ (lit. ‘[the] top of the house’).
When the modifying element is not definite (i.e. not preceded by a determiner) some of these expressions can behave like compounds (indicated in the examples immediately below with hyphens) in certain contexts, as can be seen by the final position of the modifying adjective and by the initial position of the determiner in the last example:
  • bean-taighe òg ‘a young housewife’
  • comhairle-baile ùr ‘a new town council’
  • mullach-taighe ùr ‘a new roof’
  • mo bhràthair-màthar eile ‘my other maternal uncle’.
Noun modifiers can be pluralised in plural syntactic phrases but not in plural compounds:
  • (phrase) mullaichean bheanntan ‘mountaintops’
  • (compound) cèisean-litreach ‘envelopes’
  • *cèisean-litrichean.
Genitive case nouns act as modifiers in compounds, but the genitive can be optional or archaic / inherited in syntactic phrases:
  • (compounds) sgian-arain ‘bread knife’, taigh-bìdh ‘restaurant’
  • (phrases) (optional genitive) glainne fìon(a) ‘a glass of wine’, (optional genitive) blas ara(i)n ‘taste of bread’ vs. (archaic / inherited) craobh dharaich ‘oak tree’.
The meaning of a compound can be different from the meanings of its constituent parts, at least partially, e.g.
  • taigh-beag ‘toilet’ rather than ‘small house’
  • tìr-mòr ‘mainland’ rather than ‘great / big land’
  • madadh-ruadh ‘fox’ rather than ‘red dog’
  • bean-ghlùine ‘midwife’ rather than ‘knee woman’.
In certain compounds, the head element cannot act as hypernym in subsequent references to the compound word. (A hypernym is a word denoting a category of more specific words, e.g. cutlery (hypernym) vs. knife (specific type of cutlery).) For example, taigh-beag ‘a toilet’ cannot be referred to as taigh ‘a house’. This is also true for taigh-bìdh ‘restaurant’ and taigh-seinnse ‘pub’. However, although the likes of cèis-litreach ‘envelope’ and sgian-arain ‘bread knife’ are compounds too, they can (for obvious semantic reasons) also be referred to by their hypernyms cèis and sgian respectively. Examples such as madadh-ruadh ‘fox’ fit into the first category except in dialects in which madadh on its own can refer to ‘fox’. The hypernym substitution test is therefore only partially useful in recognising compounds.

Preposed intensifiers such as fìor seem to modify phrases rarely whilst they occur normally with compounds.
  • fìor dhuine-uasal ‘a real nobleman’
  • (?) fìor bhlas ara(i)n ‘real taste of bread’.
Many expressions are ambiguous and could be interpreted either as compounds or phrases in various contexts, e.g. fear taighe ‘man of the house’, piuthar athar ‘paternal aunt’, craobh dharaich ‘oak tree’.
 
* * * * *
Generally speaking, it makes sense to hyphenate expressions that meet the above criteria for compounds, even though some might not meet all of them.

The right-branching structures discussed above can be described not as forming two clearly distinct groups of expressions, but rather as a continuum or cline, which makes the issue of hyphenation here quite challenging.

On one side, there are syntactic phrases, which function as groups of separate lexemes, both morpho-syntactically and semantically. They should be spelt as two or more separate words:
  • taigh cloiche ‘a stone house’
  • taigh mòr cloiche ‘a big stone house’
  • glainne uisge ‘a glass of water’
  • glainne mhòr uisge ‘a big glass of water’.
On the other side, there are compounds, which function as single lexemes in every respect and should be hyphenated:
  • taigh-beag ‘toilet’
  • taigh-beag glan ‘a clean toilet’.
These expressions tend to have a more weakly stressed head element, which can lead to phonetic reduction; their two parts cannot be separated by determiners and adjectives, and their meaning does not always follow from the meanings of their constituent parts, cf. taigh-beag ‘toilet’, lit. ‘small house’, etc. Certain words that are found in compounds are rarely used as independent lexemes, e.g. luchd ‘people’, as in luchd-lagha ‘lawyers’ (although luchd na Gàidhlig / Beurla is used); seinnse lit. ‘change’, as in taigh-seinnse. Many such expressions are well-established and ‘inherited’ as older formations in the language.

Certain right-branching expressions are ambiguous. They behave like phrases in that they can be divided either by (some) determiners or by adjectives, but they act as compounds in other respects or contexts:
  • bràthair d’ athar ‘your paternal uncle’ (lit. ‘your father’s brother’)
  • bràthair-athar eile dha ‘another paternal uncle of his’
  • am bràthair-athar eile ‘the other uncle’ (not *the brother of another father)
  • fear an taighe ‘the master of ceremonies, the man of the house’
  • fear-taighe òg ‘a young master of ceremonies, a young man of the house’.
It might make sense to hyphenate expressions like these when they seem to act as compounds and write them as two separate words otherwise. Alternatively, it might be recommended that a hyphen should be used only where the meaning is different to the constituent parts, e.g. taigh-beag ‘toilet’, tìr-mòr ‘mainland’.

A more radical, but also more practical solution would be to follow the Irish practice of writing all these and similar right-branching expressions as two separate words, without a hyphen.

We would be very interested to receive comments, feedback and suggestions about hyphenation in Gaelic. We are keen to understand if there is appetite for change and regularisation and clearer guidance on the use of hyphens in Gaelic. Would the strictly linguistically based recommendations outlined here be acceptable?

Please email us at mail@dasg.ac.uk.

Dr David Mandić
Professor Rob Ó Maolalaigh
Oilthigh Ghlaschu / University of Glasgow
 
 
 
Your comment has been submitted for moderation
There are no comments for this post