Principles
Microsyntactic annotations are relevant when macrosyntax segmentation isn’t precise enough to identify boundaries, this segmentation is based on syntactic dependency relations (Feuillard 1989, Rossi-Gensane 2017)
Tagset
P = Pivots
A = Actualizers
Da, Db, Dc… = Dependent segments, which are on the same level – lower case letters signal (in alphabetical order) the presence of several dependent segments
Dprop = Dependent propositional sequences, which are characterized by verbal forms
R = Relational Markers (relationneurs), i.e. Coordinative, Subordinate Conjunctions and Prepositions (labels according to traditional grammars, at the microsyntactic level) – these markers normally precede pivots, they are placed at the beginning of microsyntactic sequences
DM = Discourse Markers (introducteurs, connecteurs, ponctuants, adverbes de phrase et de commentaire…)
Categories
1. Verbal sequences are annotated as PIVOTS, even when there is a complex verbal form (presence of auxiliaries, phrasal verbs, periphrastic forms such as “être en train de + infinitive”, “aller + infinitive”, “venir + infinitive”).
| la | petite | fille | a | acheté | de | belles | chaussures |
| D(A) | D(A) | A | P | D(D(P)a) | D(D(P)a) | D(P)a | |
| et | un | très | joli | chapeau |
| R | D(D(P)b) | D(D(D(P)b)) | D(D(P)b) | D(P)b |
The parentheses (i.e. the number of parentheses) indicate syntactic complexity levels quite effectively, which are also indicated by subsequent tiers (for the analysis of dependent propositional sequences).
2. A Dprop is annotated as such in the corresponding tier (e.g. « Syntax »), a supplementary tier (e.g. « Syntax 1 ») presents the microsyntactic annotation of the Dprop sequence (« que tu viendras », in the example above).
| je | pense | que | tu | viendras | ||
A |
P | Dprop | ||||
|
R | A | P | |||
3. Annotation of relational markers – the case of coordinative markers (independent forms, which can be mentioned in a row)
| il | a | acheté | ça | et | puis | ça |
| A | P | Da | R | R | Db | |
Annotation of relational markers – the case of multi-word sequences (considered as a whole)
| vu | que | c’ | est | toujours | compliqué |
| R | A | R | D(P) | P | |
In the example above, the be form (« est ») is an empty copula (copule vide, according to Martinet) and signals the predicative status of the adjectival form « compliqué », which can be considered as the pivot of the syntactic sequence.
4. Equivalence relations (coindexicality??) between pronominal and lexical forms are marked by an apex (A prime « A’ »).
le |
problème | c’ | est | que | ce | week-end | ça | va | être | compliqué | ||
D(A’) |
A’ | A | R | |||||||||
|
R | D(D(P)) | D(P) | A | R | P | ||||||
| pseudo-cleft | ||||||||||||
Specific syntactic configurations such as clefts, pseudo-clefs constructions are mentioned in a separate tier (see examples above and below).
| c’ | est | pour | ça | que | je | t’ | appelais | ||
A |
R | R | P | A’prop | |||||
|
R | A | D(P) | P | |||||
cleft |
|||||||||
5. Disfluencies are annotated as microsyntactic categories plus the mention « disf » in subscript. If the segment cannot be analyzed as a word form (cut words, French amorces), it will be annotated as « disf » (see example below). If there are repetitions of the same words, the first forms are considered as disfluencies.
| tu | peux | tu | peux | v- | venir | avec | nous |
| Adisf | Pdisf | A | P | disf | P(P) | R | D(P(P)) |
The following segments are annotated as disfluencies (« disf »), except when a special status can be recognized (via visual or contextual information, for example…): « euh », « h. », « h », « pf ».
Disfluencies within a compound form can be annotated according their syntactic categories:
| à | la | so- | à | à | à | à | la | fin |
| Rdisf | Ddisf | disf | Rdisf | Rdisf | Rdisf |
Link to guidelines (on translation)