====== Roots ======

In what follows we only discuss how the term ‘root’ is used in the annotation oh the database. \\
That is, the cell ‘Root’ can be used as an indicator which verbs belong to the same word family. Crucially, we do not claim that the annotated form is the exact phonological form of  the root of the verb. Rather, this should be taken as a cell that can help the users group together verbs that are related by some common core (the chosen form could in principle be replaced by a random number/index). 

Still, we tried to capture what we believe is the common core of all the related verbs. The material in the cell ‘Root’ therefore refers to the verb without prefixes, thematic vowels or verbal affixes. 

Note that verbs that display suppletion between the perfective and imperfective form are annotated as having different roots. In cases where the suppletive form is related to another existing verb (//oditi// ‘to go away.pfv’, //odhajati// ‘to go away.ipfv’ where //haj// has the same root as //hoditi// ‘to walk’) the suppletive form and the related verb are marked as having the same root (in this case //h-d//, see below for hyphen).

^ Language ^ Example   ^ Root ^ Gloss & Notes   ^
| Slo/BCS  | pisati | pis |‘to write.ipfv’; the verb has no prefix, removed are the TV a and the infinitival ending |
| Slo/BCS  | napisati | pis |‘to write.pfv’; removed are the prefix //na//-, the TV //a// and the infinitival ending |
|BCS| prepisivati | pis | ‘to write.ipfv’; removed are the prefix //pre//- and the verbal suffix -//iv//, the TV //a// and the infinitival ending|
|Slo| prepisovati | pis | ‘to write.ipfv’; removed are the prefix //pre//- and the verbal suffix //-ov//, the TV //a// and the infinitival ending|


As evident from the table above, the meaning of verbs was taken as a motivating factor to determine that verbs with homophonous material belong in the same word family. In addition, another motivation for the choice of the root were alternation patterns. If multiple verbs have the same alternation pattern in their paradigm and in the aspectual pair, we treated them as having the same root. 

^ Language ^ Example   ^ Root ^ Gloss  ^
|BCS | odabrati | b-r | ‘to chosse.pfv’| 
| BCS | odabirati | b-r | ‘to chose.ipfv’ |
| BCS | sabrati | b-r |‘to sum.pfv’ |
| BCS | sabirati | b-r |‘to sum.ipfv’|
| Slo | razložiti |l-g | ‘to explain.pfv’, ‘to put down.pfv’ |
| Slo | razlagati | l-g | ‘to explain.ipfv’, ‘to put down.ipfv’ |
|Slo | položiti |l-g | ‘to lay down.pfv’ |
| Slo | polagati | l-g |‘to lay down.ipfv’ |


If the verb displays root allomorphy, we choose one of the forms of the root. While we gravitated towards what plausibly is the underlying form, this is not necessarily always so. Crucially, the same form of the root is then used with all the verbs within the same word family. 


^ Language ^ Example   ^ Root ^ Gloss  ^
| Slo |počiti | pok | ‘to burst.pfv’ |
|Slo | pokati | pok | ‘to burst.ipfv’|
| Slo | napočiti | pok | ‘to arrive.pfv’|
| BCS | izbeći | beg |‘to avoid.pfv’|
| BCS |izbegnuti |beg |‘to avoid.pfv’|
| BCS |izbegavati |beg | ‘to avoid.ipfv’|
| BCS | bežati | beg |‘to run away.ipfv’|


If there are two potential homophonous roots and only one of them displays an alternation, we use this to separate it from the other root. We return to the hyphen.


^ Language ^ Example   ^ Root ^ Gloss   ^
| BCS | primaknuti | m-k | ‘to get closer.pfv’|
| BCS | primicati | m-k | ‘to get closer.ipfv’|
| BCS | umočiti | m-č | ‘to dip.pfv’ |
| BCS | umakati | m-č |‘to dip.ipfv’|
| Slo |dotakniti | t-k |‘to touch.pfv’|
| Slo | dotikati | t-k | ‘to touch.ipfv’| 
| Slo | točiti | t-č | ‘to pour.ipfv’|
| Slo | natakati | t-č |‘to pour.ipfv’|


===== Notes about annotation =====
 
**Numbers**\\
Where two groups of verbs seem to have homophonous roots that cannot be connected with a similar meaning, we differentiate roots with numbers. 

^ Language ^ Example   ^ Root ^ Gloss    ^
| Slo | prati | p-r | ‘to wash.ipfv’|
| Slo | sprati | p-r | ‘to wash away.pfv’| 
| Slo | spirati | p-r | ‘to wash away.ipfv’ |
| Slo | odpreti | p-r1 | ‘to open.pfv’| 
| Slo | odpirati | p-r1 | ‘to open.ipfv’| 
| BCS | biti | bi | ‘to be.ipfv’ |
| BCS | zbivati | bi | ‘to happen.ipfv’ | 
| BCS |ubiti | bi2 | ‘to kill.pfv’ |
| BCS | ubijati | bi2 | ‘to kill.pfv’ |


**Hyphen** \\
We use a hyphen to indicate roots that display apophony (i.e. don’t have stable vowels), for example between the imperfective and perfective form of the verb. 

^ Language ^ Example   ^ Root ^ Gloss    ^
| BCS | ubosti | b-d | ‘to stab.ipfv’ |
| BCS | ubadati | b-d |‘to stab.ipfv’|
| BCS | bosti | b-d |‘to stab.ipfv’|
| Slo | brati | b-r | ‘to read.ipfv’ |
| Slo | prebrati | b-r | ‘to read.pfv’| 
| Slo | prebirati | b-r | ‘to read.pfv’| 



**Plus** \\
While we remove any strictly verbal material (prefixes, TV, suffixes) from the verb when annotating the ‘root’,multifunctional affixes are included in this column and are marked with a plus (+). Plus is also used in compounds (these are also annotated with 1 in the column Compounds).  

^ Language ^ Example   ^ Root ^ Gloss & Notes   ^
| Slo | beračiti | ber+ač | ‘to beg.ipfv’; berač ‘beggar’, where -//ač// is a common normalizer |
| Slo | krvaveti | krv+av | ‘to bleed.ipfv’; krvav ‘bloody’ where -//av// is a common adjectiviser |
| Slo | telovaditi | tel+o+vad | ‘to exercise.ipfv’;  //tel+o+vadba// ‘exercise’ from //tel-o// ‘body’ + //vad-b-a// ‘training’|
| BCS | krvariti | krv+ar | ‘to bleed.ipfv’|
| BCS | besniti |bes+n |‘to rage.ipfv’ |
| BCS | obelodaniti | bel+o+dan |‘to disclose.pfv’|

**Colon**\\
In the BCMS Database, the colon mark is used in cases where the annotator has the intuition there is a prefix, but the root reconstructed on this analysis does not appear in other words. e.g. //u:zeti// ‘take’, where //zeti// does not appear in any other verb. 

===== Related topic =====
[[ra|Root Allomorphy]]