The S/S Analysis

  • Peter Lasersohn
Part of the Studies in Linguistics and Philosophy book series (SLAP, volume 55)


Fortunately, the Relational/S Analysis is not the only altemative to the NP/S Analysis. Another option is the analysis I will call the S/S Analysis. Unfortunately, the S/S Analysis runs into some of the same problems as the Relational/S Analysis, plus some of its own.


Noun Phrase Thematic Role Soccer Game Official Version Sentential Conjunction 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Harris (1751) claims support for the S/S Analysis from Apollonius Dyscolus and even Aristotle, but in both cases this seems questionable. In Apollonius’ On Conjunctions,all the examples apparently involve overt sentential conjunction rather than phrasal conjunction. Apollonius does discuss phrasal conjunction in his Syntax,but there seems to favor an analysis in which phrasal conjuncts “share” elements from the surrounding context rather than one in which entire clauses are conjoined; see pp. 106–107 of Householder’s translation. Aristotle’s work was discussed above; I hope it is clear that he favored something more like the NP/S Analysis. Latham (1847), quoted by Stoddart (1849, p. 200), claimed that the S/S Analysis was “nearly coeval with the study of grammar.” However Stoddart is able to trace the analysis only as far back as the fifteenth century grammarian Aldus Manutius.Google Scholar
  2. 2.
    Francisco Sanchez de las Brozas.Google Scholar
  3. 3.
    A good summary of this debate may be found in Michael (1970, pp. 451–453). Rushton (1869, pp. 240–243) summarizes the later parts of the debate from a closer historical perspective. Dik (1968, Ch. 6) and Wierzbicka ( 1980, Ch. 7) also provides some interesting historical discussion, though with a somewhat different emphasis.Google Scholar
  4. 4.
    This observation was not original with Sanctius, though I cannot say who first made it. Linacre (ca. 1523) gives a virtually identical example sentence — after claiming in the main body of his text that conjunctions must always join identical cases, he notes in a supplement: Emi equum centum denarijs et pluris. Why hath not the cóiunction copulatiue here lyke cases? For somtyme the proprete of wordes let copulatiues and disiunctiues to couple cases: as Emi equum centuum nummis et pluris. Fui Romae et Thebis. These examples are repeated and discussed in more detail in Linacre (1524).Google Scholar
  5. 5.
    See, e.g., Breva-Claramonte (1980), Padley (1985, p. 273) or R. Lakoff (1975) for justification of a multistratal interpretation of Sanctius.Google Scholar
  6. 6.
    Roger Bacon’s Sumule Dialectices (ca. 1250), for instance, gives a variety of examples, including the traditional Two and three are five,as well as An animal is body and spirit, Morning and evening are a day,and Sor and Plato are pulling a boat The body and spirit example, like Aristotle’s two and three example, shows up repeatedly in the medieval literature.Google Scholar
  7. 7.
    Perizonius’ footnotes were first added in a 1687 edition (published by Strickius, Franeker, the Netherlands), and appear in most later editions — though not in the most recent ones. Shaw (1687, p. 158) provides a contemporaneous criticism of the S/S Analysis; however his objection is not so much against the theoretical claim that conjunctions always link whole sentences, as against the use of this sort of analysis in second language pedagogy: “If we say, That Conjunctions do only joyn Sentences together, and not properly Words, we must always allow an Ellipsis; which is hard to be understood by Children.”Google Scholar
  8. 8.
    Essentially this same example, sometimes attributed to Perizonius, actually appears at the very end of a short tract entitled Grammatica Latina,appended to Sanctius’ book Minerva (which contained the main discussion of the issue at hand).Google Scholar
  9. 9.
    Lord Monboddo.Google Scholar
  10. 10.
    The reference here is apparently to Vossius ( 1662, Book III, Chapter 3) where a conjunction is characterized as “that which conjoins a sentence to a sentence.” However, Vossius clearly intends this as contrasting not with an analysis where conjunctions can conjoin individual words, but rather with an analysis where certain conjunctions “disjoin” sentences rather than conjoining them. (Vossius argues that this latter analysis confuses form and meaning; a conjunction such as or may disjoin the MEANINGS of two sentences, but still conjoins the sentences themselves — which is just to say, it links them together syntactically.)Google Scholar
  11. 11.
    Ruddiman (1714, p. 99) does repeat the traditional rule that (certain) conjunctions must couple similar cases, and also notes (p. 78) that singular nouns joined by a conjunction require plural agreement. However, in his definition of conjunctions (p. 68), he treats them as joining whole sentences together.Google Scholar
  12. 12.
    Which is not to suggest that Home Tooke was a kind of proto-Generative-Semanticist. His reasons for classifying conjunctions as verbs were hardly the same as in Generative Semantics, and involved a severe mixing of synchronic and diachronic analysis, as well as some rather fanciful etymologies.Google Scholar
  13. 13.
    The passage in question is not included in the first (1795) edition of Murray’s English Grammar The (1802) version cited here is already the third American edition; I am not sure at exactly what point the relevant discussion was added.Google Scholar
  14. 14.
    Predicate“ here should be understood in something closer to the old Aristotelian sense than the modern logician’s sense; arguments of the verb other than the subject form part of the predicate.Google Scholar
  15. 15.
    Wierzbicka (1972) — perhaps more familiar to some readers as a manuscript dated 1967 — is a notable exception to this generalization, though she does not review the debate in any detail. Wierzbicka (1980, pp. 223–224) quotes the same passage from Beattie (1788) reproduced above, and complains of twentieth-century linguists’ “apparent unawareness that they were discovering charted territory.” Although I have some sympathy with this sort of complaint, the volume of grammatical and logical literature which has been produced over the centuries is so large that there is really no way a modern linguist can be sure of not repeating earlier ideas, except in the context of the most highly theory-internal discussions.Google Scholar
  16. 16.
    I have slightly altered some of these examples to simplify the discussion.Google Scholar
  17. 17.
    Schein uses `INFL’ for whatever thematic role is assigned to the subject argument, which in this case I have glossed, perhaps not entirely accurately, as `AGENT’. In the course of the paper, Schein makes a number of suggestions which require modifications to this formula; the final version is discussed below.Google Scholar
  18. 18.
    This logical form was provided by Schein (personal communication), and does not appear in the paper itself.Google Scholar
  19. 19.
    Schein ( 1994, Ch. 6, Sec. 2.2) suggests that at some level, outscoring is a relation between different “sides.” An individual x stands in relation 01 to an event e iff e is an outscoring event, e has two sides, and x plays for the “winning” side in e. (Similarly for 02 and the “losing” side.) Note that the “winning” and “losing” sides need not correspond to the winning and losing teams in the game, since the children may outscore the adults in the game even if each team contains a mix of children and adults. This solution necessitates a claim that an outscoring may only have other outscorings as parts, and that the parts of an outscoring must have the same winning and losing sides as the larger outscoring itself.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 1995

Authors and Affiliations

  • Peter Lasersohn
    • 1
  1. 1.Department of LinguisticsUniversity of RochesterUSA

Personalised recommendations