{
"paper_id": "W89-0102",
"header": {
"generated_with": "S2ORC 1.0.0",
"date_generated": "2023-01-19T03:44:33.253197Z"
},
"title": "A Swedish Clause Grammar and Its Implementation",
"authors": [
{
"first": "Eva",
"middle": [],
"last": "Ejerhed",
"suffix": "",
"affiliation": {},
"email": ""
},
{
"first": "A",
"middle": [],
"last": "Bstract",
"suffix": "",
"affiliation": {},
"email": ""
}
],
"year": "",
"venue": null,
"identifiers": {},
"abstract": "The paper is concerned with the notion of clause as a basic, minimal unit for the segmentation and processing of natural language. The first part of the paper surveys various criteria for clausehood that have been proposed in theoretical linguistics and computational linguistics, cind pro poses that a clause in English or Swedish or any other natural language can be defined in structural terms at the surface level as a regular expression of syntactic categories, equivalently, as a set of sequences of word classes, a possibility which has been explicitly denied by Harris (1968) and later transformational grammarians. The second part of the paper presents a grammar for Swedish clauses, and a newspaper text segmented into clauses by an experimental clause parser intended for a speech synthesis applicar tion. The third part o f the paper presents some phonetic data concerning the distribution of",
"pdf_parse": {
"paper_id": "W89-0102",
"_pdf_hash": "",
"abstract": [
{
"text": "The paper is concerned with the notion of clause as a basic, minimal unit for the segmentation and processing of natural language. The first part of the paper surveys various criteria for clausehood that have been proposed in theoretical linguistics and computational linguistics, cind pro poses that a clause in English or Swedish or any other natural language can be defined in structural terms at the surface level as a regular expression of syntactic categories, equivalently, as a set of sequences of word classes, a possibility which has been explicitly denied by Harris (1968) and later transformational grammarians. The second part of the paper presents a grammar for Swedish clauses, and a newspaper text segmented into clauses by an experimental clause parser intended for a speech synthesis applicar tion. The third part o f the paper presents some phonetic data concerning the distribution of",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Abstract",
"sec_num": null
}
],
"body_text": [
{
"text": "In tr a d itio n a l g ra m m a r a cla u se is In cu rre n t tr a n s fo r m a tio n a l g ra m m a r as p resen ted b y R a d fo r d (1 9 8 8 ), th ree ty p e s o f cla u ses a re re c o g n iz e d (see (1 ) ). per se n te n ce ), sen ten ce c o m p le x ity (p r e d ic a tio n s p e r s e n te n c e ), a n d m e a n p re d ica tio n len g th (w o rd s p e r p r e d ic a tio n ).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "W hat is a Clause in Linguistic Theory?",
"sec_num": "1"
},
{
"text": "T a b le 2 b e lo w sh ow s th a t w h erea s sen te n ce len g th varies a g re a T a b le 3 b e lo w sh o w s h o w th e fin ite (F ) a n d n o n -fin ite (N F ) p r e d ic a tio n s w ere d istrib u ted in th e gen res o f in fo rm a tiv e a n d im a g in a tiv e p rose. . . . in E n g lish a w /^cla u se c a n b e aw ay fro m its n ou n (u su a lly i f n o o th e r n o u n in te rv en e s):",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "W hat is a Clause in Linguistic Theory?",
"sec_num": "1"
},
{
"text": "F in a lly th e m a n a rriv e d w h o m th e y h ad all c o m e to m eet.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Group",
"sec_num": null
},
{
"text": "In d e s c r ib in g se n te n ces, o n e ca n still sa y th a t th ere is a c o n s titu e n t, even nd one can specify that it can be at a distance primarily if no noun intervenes.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Group",
"sec_num": null
},
{
"text": "To the extent that such problems did not arise, it would be possible to define sentence forms as short sequences of morpheme classes (or word classes), each class being expandable by a certain neighborhood of other classes (my emphasis EE). \u2022 COMP is the category o f subordinating conjunctions, including a tt as infini tive marker.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Group",
"sec_num": null
},
{
"text": "\u2022 SADV is the category o f sentence adverbs, in te, ofta, aldrig.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Group",
"sec_num": null
},
{
"text": "\u2022 ADVP is the category o f adverbial phrases.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Group",
"sec_num": null
},
{
"text": "\u2022 PREP is the category o f prepositions. (S tr a n g e r t a n d Z h i 1989, S tra n g ert 1 9 8 9 ) a n d th e o th e r b y D ie te r H u b e r (1 9 8 8 ).",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Group",
"sec_num": null
},
{
"text": "S tr a n g e r t's research p r o je c t , w h ich is still g o in g o n , stu d ies p e r c e iv e d pauses defines as clau ses \" units o f lin g u istic o rg a n is a tio n sm a lle r th a n th e se n te n ce a n d con sistin g o f a t least o n e s u b je c t a n d o n e fin ite v e r b \" . ",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Group",
"sec_num": null
},
{
"text": "T h e te x t is d iv id e d in to p a ra g ra p h s b y co n s e c u tiv e n u m b erin g . T h e p a ra g ra p h s a re d iv id e d in to o r th o g r a p h ic sen te n ces b y sen ten ce fin al p u n ctu a tio n m arks.",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Text A1",
"sec_num": null
},
{
"text": "T h e se n te n ces a re ",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "Text A1",
"sec_num": null
},
{
"text": "Proceedings of NODALIDA 1989",
"cite_spans": [],
"ref_spans": [],
"eq_spans": [],
"section": "",
"sec_num": null
}
],
"back_matter": [],
"bib_entries": {
"BIBREF0": {
"ref_id": "b0",
"title": "Nusvensk frekvensordbok baserad p\u00e5 tidningstext. 1. Graford, homografkomponenter. Data linguistica, 1. Almqvist ii Wiksell",
"authors": [
{
"first": "S",
"middle": [],
"last": "All\u00e9n",
"suffix": ""
}
],
"year": 1970,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "All\u00e9n, S. 1970. Nusvensk frekvensordbok baserad p\u00e5 tidningstext. 1. Graford, homo- grafkomponenter. Data linguistica, 1. Almqvist ii Wiksell, Stockholm.",
"links": null
},
"BIBREF1": {
"ref_id": "b1",
"title": "An experiment with heuristic parsing of Swedish",
"authors": [
{
"first": "B",
"middle": [],
"last": "Brodda",
"suffix": ""
}
],
"year": 1983,
"venue": "Proceedings o f the First Conference o f the European Chapter o f the Association fo r Computational Linguistics",
"volume": "",
"issue": "",
"pages": "66--73",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Brodda, B. 1983. An experiment with heuristic parsing of Swedish. Proceedings o f the First Conference o f the European Chapter o f the Association fo r Computational Linguistics:66-73. Pisa.",
"links": null
},
"BIBREF2": {
"ref_id": "b2",
"title": "A stochastic parts program and Noun Phrase parser for unre stricted text",
"authors": [
{
"first": "K",
"middle": [
"W"
],
"last": "Church",
"suffix": ""
}
],
"year": 1988,
"venue": "Proceedings o f the Second Conference on Applied Natural Language Processing",
"volume": "",
"issue": "",
"pages": "136--143",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Church, K .W . 1988. A stochastic parts program and Noun Phrase parser for unre stricted text. Proceedings o f the Second Conference on Applied Natural Language Processing:136-143. Association for Computational Linguistics, Austin, Texas.",
"links": null
},
"BIBREF3": {
"ref_id": "b3",
"title": "Grammatical category disambiguation by statistical optimization",
"authors": [
{
"first": "S",
"middle": [
"J"
],
"last": "Derose",
"suffix": ""
}
],
"year": 1988,
"venue": "Computational Linguistics",
"volume": "14",
"issue": "",
"pages": "31--39",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "DeRose, S.J. 1988. Grammatical category disambiguation by statistical optimization. Computational Linguistics, 14(l):31-39.",
"links": null
},
"BIBREF4": {
"ref_id": "b4",
"title": "A probability model for computer aided word class determina^ tion",
"authors": [
{
"first": "M",
"middle": [],
"last": "Eeg-Olofsson",
"suffix": ""
}
],
"year": 1985,
"venue": "A LL C Journal",
"volume": "5",
"issue": "1 & 2",
"pages": "25--30",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Eeg-Olofsson, M. 1985. A probability model for computer aided word class determina^ tion. A LL C Journal, 5(1 & 2):25-30.",
"links": null
},
"BIBREF5": {
"ref_id": "b5",
"title": "A finite state parser for Swedish with morphological analyzer and semantics",
"authors": [
{
"first": "E",
"middle": [],
"last": "Ejerhed",
"suffix": ""
}
],
"year": 1986,
"venue": "Proceedings of SAIS-86. Institutionen f\u00f6r Datavetenskap, Link\u00f6pings universitet",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Ejerhed, E. 1986. A finite state parser for Swedish with morphological analyzer and semantics. Proceedings of SAIS-86. Institutionen f\u00f6r Datavetenskap, Link\u00f6pings universitet.",
"links": null
},
"BIBREF6": {
"ref_id": "b6",
"title": "Finding clauses in unrestricted text by finitary and stochastic meth ods",
"authors": [
{
"first": "E",
"middle": [],
"last": "Ejerhed",
"suffix": ""
}
],
"year": 1988,
"venue": "Proceedings o f the Second Conference on Applied Natural Language Processmp",
"volume": "",
"issue": "",
"pages": "219--227",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Ejerhed, E. 1988. Finding clauses in unrestricted text by finitary and stochastic meth ods. Proceedings o f the Second Conference on Applied Natural Language Process- mp:219-227. Association for Computational Linguistics, Austin, Texas.",
"links": null
},
"BIBREF7": {
"ref_id": "b7",
"title": "A self-extending lexicon: description of a word learning program",
"authors": [
{
"first": "E",
"middle": [],
"last": "Ejerhed",
"suffix": ""
},
{
"first": "H",
"middle": [
"J"
],
"last": "Bromley",
"suffix": ""
}
],
"year": 1986,
"venue": "Papers from the Fifth Scandinavian Con ference o f Computational Linguistics",
"volume": "",
"issue": "",
"pages": "9--72",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Ejerhed, E. and H.J. Bromley. 1986. A self-extending lexicon: description of a word learning program. F. Karlsson [Ed.] Papers from the Fifth Scandinavian Con ference o f Computational Linguistics:b9-72. Publication No. 15, Department of General Linguistics, University of Helsinki.",
"links": null
},
"BIBREF8": {
"ref_id": "b8",
"title": "Finite state parsing",
"authors": [
{
"first": "E",
"middle": [],
"last": "Ejerhed",
"suffix": ""
},
{
"first": "K",
"middle": [
"W"
],
"last": "Church",
"suffix": ""
}
],
"year": 1983,
"venue": "Papers from the Seventh Scandinavian Conference o f Linguistics",
"volume": "",
"issue": "",
"pages": "410--432",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Ejerhed, E. and K .W . Church. 1983. Finite state parsing. F. Karlsson [Ed.j. Papers from the Seventh Scandinavian Conference o f Linguistics:410-432. Publication No. 9, Department o f General Linguisics, University of Helsinki.",
"links": null
},
"BIBREF9": {
"ref_id": "b9",
"title": "The process of language understanding: A few issues in contemporary psycholinguistics",
"authors": [
{
"first": "G",
"middle": [
"B"
],
"last": "Flores D 'arcais",
"suffix": ""
},
{
"first": "R",
"middle": [],
"last": "Schreuder",
"suffix": ""
}
],
"year": 1983,
"venue": "The Process o f Language U n",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Flores d 'Arcais, G.B. and R. Schreuder. 1983. The process of language understanding: A few issues in contemporary psycholinguistics. G.B. Flores d 'Arcais and R. Jarvella [Eds.]. The Process o f Language U n d e r s t a n d i n g Wiley, Chichester.",
"links": null
},
"BIBREF10": {
"ref_id": "b10",
"title": "Mathematical structures o f language",
"authors": [
{
"first": "Z",
"middle": [],
"last": "Harris",
"suffix": ""
}
],
"year": 1968,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Harris, Z. 1968. Mathematical structures o f language. Wiley, New York.",
"links": null
},
"BIBREF11": {
"ref_id": "b11",
"title": "Aspects o f the communicative function of voice in text intonation-Constancy and variability in Swedish fundamental frequency contours. Depart ment of Computational Linguistics",
"authors": [
{
"first": "D",
"middle": [],
"last": "Huber",
"suffix": ""
}
],
"year": 1988,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Huber, D. 1988. Aspects o f the communicative function of voice in text intonation- Constancy and variability in Swedish fundamental frequency contours. Depart ment of Computational Linguistics, University of G\u00f6teborg, Department of Infor mation Theory, Chalmers Institute of Technology, G\u00f6teborg, and Department of Linguistics and Phonetics, University of Lund.",
"links": null
},
"BIBREF12": {
"ref_id": "b12",
"title": "The resolution of morphological ambiguities",
"authors": [
{
"first": "F",
"middle": [],
"last": "Karlsson",
"suffix": ""
}
],
"year": 1989,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Karlsson, F. 1989. The resolution of morphological ambiguities. Paper presented at the Scandinavian Conference of Computational Linguistics, Reykjavik, June 26-28, 1989.",
"links": null
},
"BIBREF13": {
"ref_id": "b13",
"title": "Computational analysis of predicational structures in English",
"authors": [
{
"first": "H",
"middle": [
"N"
],
"last": "Kucera",
"suffix": ""
}
],
"year": null,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Kucera, H. n.y. Computational analysis of predicational structures in English. Brown University, Providence, R.I. (unpublished).",
"links": null
},
"BIBREF14": {
"ref_id": "b14",
"title": "The analysis o f the English verbal group",
"authors": [
{
"first": "H",
"middle": [],
"last": "Kucera",
"suffix": ""
}
],
"year": 1985,
"venue": "ICAME Sixth International Conference on English Language Research on Com puterised Corpora",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Kucera, H. 1985. The analysis o f the English verbal group. Paper presented to the ICAME Sixth International Conference on English Language Research on Com puterised Corpora, Lund (unpublished).",
"links": null
},
"BIBREF15": {
"ref_id": "b15",
"title": "HP-systemet som genv\u00e4g vid syntaktisk m\u00e4rkning av texter",
"authors": [
{
"first": "G",
"middle": [],
"last": "K\u00e4llgren",
"suffix": ""
}
],
"year": 1984,
"venue": "Sven skans beskrivning",
"volume": "14",
"issue": "",
"pages": "39--45",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "K\u00e4llgren, G. 1984a. HP-systemet som genv\u00e4g vid syntaktisk m\u00e4rkning av texter. Sven skans beskrivning, 14:39-45, Lunds universitet.",
"links": null
},
"BIBREF16": {
"ref_id": "b16",
"title": "HP-A heuristic finite state parser based on morphology",
"authors": [
{
"first": "G",
"middle": [],
"last": "K\u00e4llgren",
"suffix": ""
}
],
"year": 1984,
"venue": "",
"volume": "",
"issue": "",
"pages": "155--162",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "K\u00e4llgren, G. 1984b. HP-A heuristic finite state parser based on morphology. A. S\u00e5gvall Hein [Ed.]. De nordiska datalingvistikdagama 1985:155-162. Centrum f\u00f6r Dator lingvistik, Uppsala Universitet.",
"links": null
},
"BIBREF17": {
"ref_id": "b17",
"title": "Transformational Grammar: A First Course",
"authors": [
{
"first": "A",
"middle": [],
"last": "Radford",
"suffix": ""
}
],
"year": 1988,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Radford, A. 1988. Transformational Grammar: A First Course. Cambridge University Press, Cambridge.",
"links": null
},
"BIBREF18": {
"ref_id": "b18",
"title": "Pause patterns in Swedish: A project presentation and some data. Fonetik-89",
"authors": [
{
"first": "E",
"middle": [],
"last": "Strangert",
"suffix": ""
},
{
"first": "M",
"middle": [],
"last": "Zhi",
"suffix": ""
}
],
"year": 1989,
"venue": "Speech Transmission Laboratory Quarterly Progress and Status Report",
"volume": "1",
"issue": "",
"pages": "27--31",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Strangert, E. and M. Zhi. 1989. Pause patterns in Swedish: A project presentation and some data. Fonetik-89. Speech Transmission Laboratory Quarterly Progress and Status Report, 1:27-31. KTH, Stockholm.",
"links": null
},
"BIBREF19": {
"ref_id": "b19",
"title": "Pauses, syntax and prosody. Paper presented to the Nordic Prosody V meeting in",
"authors": [
{
"first": "E",
"middle": [],
"last": "Strangert",
"suffix": ""
}
],
"year": 1989,
"venue": "",
"volume": "",
"issue": "",
"pages": "",
"other_ids": {},
"num": null,
"urls": [],
"raw_text": "Strangert, E. 1989. Pauses, syntax and prosody. Paper presented to the Nordic Prosody V meeting in Turku, Finland, August 23-25, 1989 (to appear).",
"links": null
}
},
"ref_entries": {
"FIGREF0": {
"num": null,
"type_str": "figure",
"text": "d efin e d as a u n it con sistin g o f a s u b je c t a n d a p r e d ic a te . T h e term s s u p p o s itu m a n d a p p o s itu m w ere used in sch o la stic g r a m m a r t o d e n o te th e sy n ttictic fu n c tio n s o f th ese tw o b a sic p a rts o f a cla u se. T r a d itio n a l g ra m m a r malces a d is tin c tio n b etw ee n m a in clau ses a n d d e p e n d e n t cla u ses.",
"uris": null
},
"FIGREF1": {
"num": null,
"type_str": "figure",
"text": "o r d in g to R^ldford (1 9 8 8 ) \" th e th ree C la u se ty p e s d iffer p rin c ip a lly in th a t O rd in a ry C lau ses co n ta in b o th I a n d C , E x c e p t io n a l cla u ses c o n ta in I (= in fin i-tival to ) b u t n o t C , a n d S m all C lau ses c o n ta in n e ith er C n o r I. M o r e o v e r , b o t h E x c e p tio n a l C lau ses a n d S m all C la u ses a re h ig h ly r e s tricte d in th eir d is trib u tion : fo r e x a m p le , E x c e p tio n a l C la u ses ty p ic a lly o c c u r o n ly as th e C o m p le m e n ts o f certa in s p e cific ty p e s o f v erb s; a n d S m a ll C la u ses o c c u r m a in ly as th e C o m p le m ents o f a su b set o f V erb s a n d P r e p o s itio n s ...\" It sh o u ld b e n o te d th a t I here is ten se, m o d a l, o r in fin itiv a l t o , a n d C is co m p le m e n tiz e r . E x a m p le s o f o r d in a r y clau ses are g iv en in (2 ), (3 ) a n d (4 ) b e lo w . r o v e th e p r o j e c t ' ro v e th e p r o je c t -In c o m p u ta tio n a l lin g u istics, th ere is n o sin gle answ er t o th e q u estio n o f w h a t a cla u se is, sin ce th is d e p e n d s o n th e p a rticu la r g ra m m a tica l th e o ry ch osen in a g iv e n c o m p u ta tio n a l fra m e w ork . In o r d e r t o illu stra te o n e p a rticu la r a n d e x p licit n o tio n o f cla u se, o r m ore p r e cis e ly p r e d ic a tio n , in c o m p u ta tio n a l lin g u istics, I w ant to q u o te an in terestin g s tu d y b y H en ry K u 5 e ra (m s, 1 985) o n th e co m p u ta tio n a l a n a lysis o f p re d ica tio n a l str u c tu r e s in th e B r o w n C o rp u s. H e c o n s id e rs a p r e d ic a tio n t o b e , first o f all, an y v e rb o r v e rb a l g r o u p w ith a ten sed v e rb th a t is s u b je c t t o c o n c o r d (fo r p e rso n an d n u m b e r) w ith its g r a m m a t ic a l s u b je c t . T h e s e v e rb a l c o n s tr u c tio n s h e calls finite p re d ica tio n s. In a d d itio n t o th a t, h e a ls o in clu d e s in his a n a lysis n o n -fin ite p re d ica tio n s, c o n sistin g o f in fin itiv a l c o m p le m e n ts , g e ru n d s a n d p a rticip le s. W h a t h e d id in his s tu d y w as to id e n tify a n d c la ss ify a ll th e p r e d ic a tio n s , w h ich w ere 145,287 in all th e 5 4 ,7 2 4 se n ten ces o f th e B r o w n C o rp u s . T a b le 1 sh ow s fo r ea ch g e n re in th e co r p u s , th e m ean sen ten ce le n g th (w o rd s",
"uris": null
},
"FIGREF2": {
"num": null,
"type_str": "figure",
"text": "t d e a l b e tw e e n a m ean o f 21 w o rd s p e r se n te n ce in in fo rm a tiv e p r o s e (I N F O ) a n d 13 w o r d s p e r sen ten ce in im a g in a tiv e p r o s e (I M A G ) , sen ten ce c o m p le x ity d o e s n o t v a ry th a t m u ch b e tw e e n gen res: 2 .8 0 versu s 2 .3 8 p re d ica tio n s p e r sen ten ce.",
"uris": null
},
"FIGREF3": {
"num": null,
"type_str": "figure",
"text": "h a t K u C era co n s id e rs as th e m a in resu lt o f his s tu d y is th e la ck o f co r r e la tio n b e tw een se n te n ce le n g th an d se n ten ce c o m p le x ity , a n d it is in d e e d su rp risin g . K uC era's s tu d y w as c o n c e r n e d w ith fin d in g , c o u n tin g a n d cla s s ify in g p re d i ca tio n s u n its (v e r b a l g r o u p s ) in th e B ro w n C o r p u s . It w as n o t c o n c e r n e d w ith w h at w ou ld h ave b een an e v en m o r e d ifficu lt g o a l, th a t o f fin d in g en tire cla u se units, in th e sense o f d e m a r c a tin g th eir b e g in n in g s a n d e n d in g s. T h e r e is a n o b v iou s rela tion b etw ee n p r e d ic a tio n s a n d cla u ses, in th a t a r e a so n a b le d e fin itio n o f clau se, I th in k , w o u ld b e o n e in w h ich th ere is o n e p r e d ic a tio n , in K u C era ' s sense o f th e te rm , p e r cla u se. In E je rh e d (1 9 8 8 ), w h ich is a c o m p u ta tio n a l lin g u istic s tu d y o f cla u ses in E n glish , d o n e in c o lla b o r a tio n w ith K e n C h u rch w h en I v is ite d A T T B e ll L a b o ra tories 1 9 8 6 -8 7 , I u sed a d e fin itio n o f cla u se th a t d iffered s o m e w h a t fr o m th e on e co n sid e re d in th e p re v io u s p a ra g ra p h . In m y d e fin itio n o f cla u se in E n g lish , Computational Linguistics -Reykjavtk 1989 o n ly fin ite a n d to -in fin itiv a l p r e d ic a tio n s are criteria ! fo r cla u s e h o o d . O th e r in fin itiv a l p r e d ic a tio n s , g eru n d s a n d p a rticip le s are n o t ta k en t o im p ly th e p resen ce o f a cla u se u n it. A n o t h e r fe a tu re o f m y d e fin itio n o f cla u se th a t w as u sed in parsin g cla u ses in u n re s tricte d t e x t, is th a t th e o p e n in g o f a n ew cla u se a lw ays im p lies th e closu re o f th e p re v io u s cla u se u n it, w h e th er o r n o t th is u n it is c o m p le te w ith s u b je c t an d p r e d ic a te , o r c o m p le te w ith r e sp e ct t o th e a rg u m en t s tru ctu re o f its p red ica te. T o illu stra te th is n o -n e s tin g o f cla u ses, th e sen ten ce in (2 ) is r e p r o d u c e d in (5 ) b e lo w w ith cla u se b o u n d a r ie s in serted w h ere th e cla u se parsers d e s crib e d in E jerh ed (1 9 8 8 ) w o u ld p la c e th em . (5) [Mary mig h t think] [that he will resign] T h e r e a re sev era l rea son s fo r th e m o v e t o a d o p t th e h y p o th e sis th a t clau ses d o n o t n est, a t a v e ry su p e rficia l level o f s y n ta c tic stru ctu re. T h e first rea son is th a t th e h y p o th e s is m ak es po.ssible an e x c e e d in g ly sim ple d e fin itio n o f, a n d r e c o g n itio n a lg o rith m fo r, clau ses: a cla u se ca n b e d e fin ed as a set o f p e rm issib le seq u en ce s o f w o r d classes b y m ean s o f a regu la r ex p ression , i.e. b y u sin g th e o p e r a tio n s o f c o n c a te n a tio n , u n ion an d K leen e sta r o n elem en ts th a t are w o r d classes. T h a t su ch a s im p le d e fin itio n o f cla u ses, o r sen ten ce fo rm s as h e ca lle d th em , w as p o s s ib le , w as s o m e th in g H arris co n s id e re d , b u t re jected in th e fo llo w in g p a ssa g e fr o m H a rris (1 9 6 8 :3 1 -3 2 ):",
"uris": null
},
"FIGREF4": {
"num": null,
"type_str": "figure",
"text": "th o u g h w ith n o n -c o n tig u o u s p a rts: th e s u b je c t a b o v e is M A N w ith a d jo in e d T H E o n th e left a n d W H O M . . . a fte r th e v e r b o n th e right.^^ B u t th e d ifficu lty lies in fo r m u la tin g a c o n s tr u c tiv e d e fin itio n o f th e sen ten ce.F o r i f w e w ish t o c o n s tr u c t th e se n te n ce b y d efin in g a s u b je c t con stitu e n t a n d th e n n e x t t o it a v e rb (o r p r e d ic a te ) co n s titu e n t, w e are u n a b le to s p e c ify th e s u b je c t i f it is d is c o n tig u o u s , b e c a u s e w e ca n n o t s p e cify th e lo c a t io n o f th e s e c o n d p a rt (th e a d ju n c t at a d is ta n c e ). A t least w e ca n n o t s p e c ify th e lo c a t io n o f th e d ista n t a d ju n c t u n til w e h ave p la ce d th e v e r b c o n s titu e n t in re sp e ct t o th e s u b je c t ; b u t w e c a n n o t p l2 u;e th e v erb in re sp e ct t o th e s u b je c t as a sin gle e n tity unless th e s u b je c t has b een fu lly specified.^\u00c2",
"uris": null
},
"FIGREF5": {
"num": null,
"type_str": "figure",
"text": "T h e se n te n ce d iscu sse d in th e p a ssa g e a b o v e w o u ld b e p a rsed as in d ica te d in (6 ), g iv e n th e cla u se g ra m m a r o f E je rh e d (1 9 8 8 ).",
"uris": null
},
"FIGREF6": {
"num": null,
"type_str": "figure",
"text": "[the m a n arrived] [whom they ha d all come to meet] T h e s e co n d rea son fo r th e h y p o th e s is th a t cla u ses d o n o t n est has t o d o w ith p e rfo rm a n ce c o n s id e r a tio n s , i.e. o b s e r v a tio n a l d a ta fr o m stu d ie s in p s y c h o lin g u istics a n d p h o n e tics.F or a rev ie w o f th e cla u sa l h y p o th e s is in p s y ch o lin g u is tics a n d stu d ie s rela t in g to it, th e rea d er is referred t o F lo re s d ' A r c a is a n d S ch re u d er (1 9 8 3 :1 4 -1 9 ).T h e y p resen t th e cla u sa l h y p o th e s is as a v ie w o f se n te n ce co m p r e h e n s io n th a t isch a ra cte riz e d b y tw o m a jo r fea tu res. F irst, cla u ses are taken t o b e th e p rim a ry u n its o f n o rm a l sp eech p e r c e p tio n . In c o m in g m a teria l is o rg a n iz e d in im m e d ia te m e m o r y cla u se b y clau se; th e listen er o r rea d e r a ccu m u la te s e v id e n c e u n til th e en d o f a clau se. S eco n d , a t th e en d o f a cla u se , w o rk in g m e m o r y is cle a re d o f su rfa ce g ra m m a tica l in fo rm a tio n a n d th e co n te n t o f th e cla u se is re p re sen ted in a m o re a b stra ct form . T h e y p o in t o u t th a t th ese tw o m a jo r p r o p e r tie s o f th e h y p oth esis axe lo g ica lly in d e p e n d e n t. P h o n e tic e v id e n ce fo r th e se g m e n ta tio n o f sp e e ch (in p e r c e p tio n as w ell as p r o d u c tio n ) a t th e level o f cla u ses, as s tr u c tu r a lly d e fin e d u n its, w ill b e d is cu ssed in th e last s e ctio n o f th e p a p e r , a fter a p re s e n ta tio n a n d illu s tra tio n o f a stru ctu ra l d efin itio n o f S w edish clau ses. 2 A Swedish Clause Grammar T h is g ra m m a r fo r S w edish cla u ses has th e sa m e s tru ctu ra l u n its as ta rg e ts as th e g ra m m a r fo r E n glish cla u ses in E je rh e d (1 9 8 8 ), m o d u lo th e d iffe re n ce b etw ee n th e tw o la n gu ages, i.e. fin ite (te n s e d ) cla u ses a n d in fin itival cla u ses in tr o d u c e d b y att are clau ses. In a d d itio n , th ere axe th ree ty p e s o f cla u se fra g m en ts : v e rb ph rase fra g m en ts, n o u n p h ra se fra g m e n ts a n d a d v e r b fra g m e n ts. In an a p p e n d ix t o th is p a p e r, th ere is a S w edish n e w s p a p e r te x t fr o m A p r il 1984 w h ich has b e e n seg m en ted in to cla u ses a n d cla u se fra g m e n ts , la b e lle d to th e righ t a c c o r d in g t o th e ty p e o f u n it in th e g ra m m a r th a t th e y in sta n tia te . T h e ca teg ories th a t are criteria ! t o th e id e n tifica tio n o f a cla u se o r cla u se fra g m e n t a c c o r d in g t o th e g ra m m a r, h ave b een la b e lle d u n d ern ea th . -comp 5. sc-coord (COORD) (PREP) COMP ... COORD (SADV) VFIN/VSUP 6PP/ADVP/SADV* A fe w w o rd s o n th e n o ta tio n u sed in th e g ra m m a r are req u ired . F o r re a d a b il ity, c o n c a te n a tio n is rep rese n te d s im p y b y ju x ta p o s itio n . U n ion (i.e . a lte rn a tiv es) is re p rese n ted b y / , a n d th e s p e c ia l ca se w h ere so m e th in g altern a tes w ith n o th in g (i.e . o p tio n a lity ) is rep resen ted b y ( ) . K le e n e sta r is represen ted b y * , w h ich h as s c o p e o v e r / . T h e th ree d o ts . . . sh o u ld b e read as a varia b le o v e r a n y w ord class. \u2022 COORD is th e c a te g o r y o f c o o r d in a tin g c o n ju n c tio n s , och, eller, m en . \u2022 NP is a n o n -r e c u r s iv e n o u n p h ra se co n sistin g o f a n y p re n o m in a l m od ifiers p lu s h ea d n o u n . NP d o e s n o t in c lu d e a n y p o s tn o m in a l m o d ifie rs. F or the c o n c e p t o f su ch a n o u n p h ra se as a p p lie d t o E n g lish , see C h u rch (1 9 8 8 ). \u2022 NP' consists o f a non-recursive NP followed by postnominal modifiers that are non-clausal, i.e. prepositional phrases PP, or adverbs ADV. Thus, NP' = NP PP/ADV* \u2022 VFIN is th e c a te g o r y o f fin ite v e rb s, a c tiv e o r p a ssiv e, an d VSUP is th e ca te g o r y o f s u p in u m fo r m s o f v erb s o c c u r r in g a fte r th e a u x ilia ry hava. B e ca u se fin ite fo r m s o f hava ca n b e o p tio n a lly d eleted in su b o rd in a te cla u ses in S w ed ish , it is n ecessa ry t o a llo w o c c u r r e n c e s o f VSUP in su ch ca ses t o cou n t as fin ite.",
"uris": null
},
"FIGREF7": {
"num": null,
"type_str": "figure",
"text": "a ch o f th e regu la r ex p ression s 1 th r o u g h 11 c o n s titu te s ein a lte rn a tiv e d ef in ition o f cla u se o r cla u se fra g m e n t. T h e w a y th a t th ese a lte rn a tiv e d e fin itio n s in te ra ct in th e p r o c e s s in g o f a te x t is v e ry im p o r ta n t. In ca ses w h ere tw o o r m o r e a lte rn a tiv e a n a lyses c o m p e t e , th e regu lar e x p r e s s io n th at m a tc h e s th e lo n g e s t su b strin g wins. T h is c a n b e illu stra ted b y c o n s id e rin g h o w th e first se n te n ce o f th e te x t in th e a p p e n d ix is p ro ce s s e d . T h e sen te n ce is re p e a te d b e lo w w ith n u m b e rs in d ica tin g lin ea r p o s itio n s in th e strin g o f w ord s. regu la r ex p ression 9, N P -fra g m e n t, m a tch e s th e s trin g o f w o r d s fr o m 0 to 1. T h e regu lar exp ressio n 7, V P -fr a g m e n t, m a tch e s th e s trin g o f w o r d s fr o m 1 to 4. T h e regu lar e x p ression 1, n on -in v e rte d m a in cla u se, m a tch e s th e s trin g o f w ord s from 0 t o 4. T h is is th e e x p ression th a t m a tch e s th e lo n g e st su b strin g , and it w in s ov e r th e a lte rn a tiv e a n a ly ses o f th e strin g fr o m 0 t o 4. T h e sta tu s o f th e im p le m e n ta tio n o f th is p a r tic u la r cla u se g ra m m a r fo r Sw edish is th a t it is in th e p ro ce s s o f b e in g im p le m e n te d . W h a t th a t m e a n s, is th a t I d o n o t y e t h a v e a ru n n in g p r o g r a m fo r S w edish th a t a u to m a tic a lly d e cid e s th e lo c a tio n o f b o u n d a rie s b etw een cla u ses a n d cla u se fra g m e n ts in u n re s tricte d te x t. T h is is an a m b itio u s a n d lo n g ra n g e g o a l, a n d th e b ig g e s t p r o b le m in d e v e lo p in g su ch a p ro g ra m is le x ica l. E a ch w o r d in a te x t has t o b e la b e lle d w ith a u n iq u e s y n ta c tic c a te g o r y (in c lu d in g in fo r m a tio n a b o u t th e fo r m o f th e w o r d ) b e fo re a n y m a tch in g a g a in st th e reg u la r ex p re ssio n s in th e g ra m m a r ca n tak e p la ce. T h e c a te g o r y la b e l a ssign ed t o a w o r d has t o b e th e o n e th a t is c o r r e c t fo r th e w o r d in its co n te x t o f o c c u r r e n c e . A su ccessfu l a p p ro a ch t o th e p r o b le m o f a u to m a tic a lly a ssign in g u n iq u e a n d c o rre ct sy n ta c tic ca teg orie s t o E n g lish w o r d s in c o n te x t is p r o b a b ilis tic (C h u r c h 1988, D e R o s e 1988, E e g -O lo fs s o n 1 9 8 5 ). T h is is o n e o f sev era l a p p r o a c h e s th a t w ill b e a p p lie d to S w edish in th e c o n te x t o f a jo in t c o r p u s b a se d resea rch p r o je c t betw een th e u n iversities o f S to c k h o lm a n d U m e \u00e5 (K \u00e4 llg r e n , E je r h e d ) th a t w ill sta rt in th e fa ll o f 1989. A n o th e r a p p r o a c h t o th e d is a m b ig u a tio n o f th e s y n ta c tic c a te g o r y a n d fo r m o f a w o rd in c o n te x t is ru le b a sed , co n stra in t b a se d o r h eu ristic, a n d th e d isa m b ig u a tio n b etw een a lte rn a tiv e an a ly ses o f a w o r d is d o n e as an in te g ra te d p a rt o f th e p a rsin g o f a te x t, ra th er th an as a sep a ra te s u b r o u tin e c o m p le te d b e fo r e p^lrsing b eg in s. A v ersion o f th is a p p r o a c h has b e e n a p p lie d t o S w edish w ith su ccessfu l (9 5 % c o r r e c t) resu lts (B r o d d a 1983, K \u00e4 llg re n 1984a, 1 9 8 4 b ).",
"uris": null
},
"FIGREF8": {
"num": null,
"type_str": "figure",
"text": "im e d in his p a p e r a t th is co n fe r e n c e , o n th e b a sis o f his recen t research o n d isa m b ig u a tio n , th a t m o re th a n 6 0 % o f th e c o n s e c u tiv e w o rd s in a S w edish te x t are at lea st tw o -w a y a m b ig u o u s , as c o m p a r e d w ith 4 5 % in E nglish a c c o r d in g t o D e R o s e (1 9 8 8 ), a n d 1 1 % in F in n ish . K a r ls s o n 's fig u re fo r S w edish tallies w ith w h a t is r e p o r te d in A ll\u00e9 n (1 9 7 0 :X V , X X V ) : 6 4 5 ,0 0 0 o u t o f th e 1,000,669 w o rd s o f th e S w edish c o r p u s P re ss-6 5 w ere h o m o g r a p h s , a n d th a t 2im ou n ts to 6 4 .5 % . W h a t I h a v e b y w a y o f im p le m e n ta tio n a t th is tim e is a m o d ific a tio n o f th e fin ite s ta te p a rser fo r S w ed ish , d e s c r ib e d in E je rh e d & C h u rch (1 9 8 3 ), E jerh ed & B r o m le y (1 9 8 5 ), a n d E je rh e d (1 9 8 6 ). S u b je c t t o th e lim ita tion s o f its le x ic o n , w h ich is c u r r e c tly b e in g e x p a n d e d , th e m o d ifie d parser, in its p a rsin g o f o r th o g r a h p ic se n ten ce s as in p u t, is c a p a b le o f id e n tify in g an d assign ing c o n stitu e n t s tr u c tu r e t o su b strin g s th a t can b e p u t in d irect co r r e s p o n d e n c e w ith th e 11 d ifferen t cla u ses a n d cla u se fra g m en ts en u m e ra ted in th e n ew cla u se g ra m m a r d e s c r ib e d here. 3 Phonetic Data concerning Clause Boundaries T h e r e a re tw o recen t p h o n e t ic stu d ie s o f s p o k e n S w edish, b a se d o n record in g s o f se v era l d ifferen t sp ea k ers re a d in g th e sa m e te x ts a lou d . O n e is b y E v a S tran gert",
"uris": null
},
"FIGREF9": {
"num": null,
"type_str": "figure",
"text": "in 2 te x ts o f a to t a l o f 8 1 0 w o rd s rea d a lo u d b y 10 differen t spea k ers a t 3 differ en t s p e e ch ra tes, a n d th e a c o u s tic a n d g ra m m a tica l p rop erties o f su ch pau ses.T h e first o f th e tw o te x ts is id e n tica l t o th e te x t in th e a p p e n d ix o f th is p a p er.A c o u s t ic a lly , a p e r c e iv e d p a u se ca n b e sign a lled in several differen t w ays: b y fi n al le n g th e n in g , a s p e c ia l fu n d a m e n ta l fre q u e n cy c o n to u r , silen ce, a n d /o r v o ic e q u a lity irreg u la ritie s. S tra n g e rt a n d Z h i (1 9 8 9 ) re p o rts fin din gs p rim a rily c o n ce r n in g th e se a c o u s tic p r o p e r tie s o f th e pa u ses p e rce iv e d b y tw o differen t ju d g e s .S tra n g e rt (1 9 8 9 ) is a ls o co n c e r n e d w ith th e d istrib u tio n o f th e p e r c e iv e d pau ses in re la tio n t o th e fo llo w in g k in ds o f b o u n d a rie s: p a ra g ra p h , se n te n ce, cla u se and p h ra se.U sin g th e d e fin itio n o f cla u se p resen te d in this p a p e r, I h a v e se g m en ted the tw o te x ts u sed in S tr a n g e r t's s tu d y a n d fo u n d th a t th e y c o n sist o f a to t a l o f 115 u n its th a t are cla u se s o r cla u se fra g m e n ts. T h e n u m b e r o f p erce iv e d pauses a t th e se 115 cla u s e b o u n d a r ie s is p rese n te d in T a b le 4 b e lo w , fo r w h ich I am in d e b te d t o E v a S tra n g ert. A p e rce iv e d p a u se is h ere a p a u se ju d g e d b y b o th o f th e tw o ju d g e s t o b e p resen t in th e sp e e ch o f a t least 5 o f th e 10 speakers.F o r th e p u r p o s e s o f th is ta b le , all cla u se b o u n d a rie s h ave b ee n in clu d ed , w h e th er th e y are se n te n ce in tern a l, o r h a p p e n t o c o in c id e w ith sen ten ce b o u n d a rie s o r p a ra g ra p h b o u n d a r ie s . In S tra n g ert (1 9 8 9 ) th ese th ree b o u n d a r y co n d itio n s are tre a te d sep a ra tely . The frequency o f clause boundaries where pauses were perceived. T h e s tu d y o f H u b e r (1 9 8 8 ) is co n c e r n e d w ith in to n a tio n u n its in r e co rd in g s o f 3 n e w sp a p e r te x ts read a lo u d b y 4 differen t sp ea k ers o f S w ed ish , a to t a l o f 2.2 h ou rs o f co n n e c te d sp e e ch . H e d efin es th e c o n c e p t o f in to n a tio n u n it in p u r e ly a co u stica l te rm s, rela te d t o fu n d a m e n ta l fr e q u e n c y o n ly , a n d d ev ises a m e t h o d o f a u to m a tic a lly se g m e n tin g co n n e c te d sp e ech in to su ch in to n a tio n u n its. T h e a d v a n ta g e o f th is se g m e n ta tio n p r o c e d u r e is th a t it m ak es n o re fe re n ce t o e ith e r h igh er level lin g u istic in fo rm a tio n c o n c e r n in g sy n ta x , o r t o lo w e r level p h y s io lo g ica l in fo r m a tio n c o n c e r n in g p a u sin g , b re a th in g , p h o n a tio n o n s e t o r offse t e tc.H e a rrives a t a to t a l o f 1664 in to n a tio n u n its in th e a c c u m u la te d te x t m a te ri al (3 te x ts , 4 sp ea k ers). T a b le 5 sh ow s th e g r a m m a tic a l co rre la te s o f th e 1664in to n a tio n u n its, a v eraged a cross fo u r spea k ers a n d th re e te x ts . F o r th e e x a c t d e fin ition s o f th e g ra m m a tica l u n its, see H u b e r (1 9 8 8 :7 8 ). O f in terest h ere is th a t h e defines as sen ten ces \" g ra p h ic sen ten ces th a t b e g in w ith a c a p ita l le tte r a n d en d w ith a fu ll s to p (o r so m e o th e r m a rk o f ' fin a l' p u n c t u a t io n )\" . A n d he",
"uris": null
},
"FIGREF10": {
"num": null,
"type_str": "figure",
"text": "Frequency o f intonation units corresponding to different grammatical categories. U n fortu n a te ly , th ese figures c a n n o t b e d ir e c t ly re la te d t o th e n o tio n s o f cla u se an d cla u se fra g m en t d iscu ssed in th is p a p e r, b e c a u s e th e d e fin itio n s o f th e g r a m m a tica l c a te g o rie s d o n o t agree. H ow ev e r, it is lik ely th a t w e ca n e q u a te m o n oclau sal s e n te n c e s (w h ich a c c o u n te d fo r 6 3 .6 % o f th e 1 -IU -p e r-s e n te n ce th a t o c cu rre d ) w ith a su b set o f m a in cla u ses (R u le s 1 -3 in th e S w edish cla u se g r a m m a r ), clau ses w ith eith er a s u b se t o f m ain cla u ses (in th e ca se o f m u lticla u sa l sen ten ce s) o r a su b set o f s u b o r d in a te cla u ses (R u le s 4 -6 ) , a n d in itia l ad verbials w ith a d v e r bfragm en t (R u le 1 1 ), a n d th ese th ree c a te g o rie s to g e th e r a c c o u n t fo r 6 0 % o f all in to n a tio n u n its. It is a lso lik ely th a t su b je c t c o r r e s p o n d s t o N P -fr a g m e n t, a n d verbphrase t o V P -fr a g m e n t o n th e b a sis o f th e illu stra tiv e e x a m p le s o f th ese ca teg o rie s in H u b e r (1 9 8 8 :8 3 -8 5 ). I f s o , c lo s e t o 7 0 % o f H u b e r 's in to n a tio n u n its w ou ld c o rre s p o n d t o a cla u se o r cla u se fra g m e n t in th e sen se o f th e p re sen t p a p e r.",
"uris": null
},
"FIGREF11": {
"num": null,
"type_str": "figure",
"text": "In o rd e r t o esta b lish th e e x a c t e x te n t t o w h ich th e n o tio n s o f cla u se a n d cla u se fra gm en t p r o p o s e d h ere c o rre la te w ith th e in to n a tio n u n its fo u n d in H u b e r ' s stu dy, a sep a ra te s tu d y is b e in g u n d erta k en in c o lla b o r a tio n w ith H u b e r.",
"uris": null
},
"FIGREF12": {
"num": null,
"type_str": "figure",
"text": "d iv id e d in to n o n -re cu rs iv e clau ses o r cla u se fra g m en ts m arked b y [ ], a n d ea ch su ch u n it is la b e lle d a c c o r d in g to th e S w edish cla u se g ra m m a r p re se n te d in th is p a p er.",
"uris": null
},
"TABREF2": {
"content": "",
"type_str": "table",
"text": "",
"num": null,
"html": null
},
"TABREF4": {
"content": "",
"type_str": "table",
"text": "",
"num": null,
"html": null
}
}
}
}