Sag, Wasow, and Bender (2003): Syntactic Theory: A formal introduction, Second Edition

Errata as of Autumn 2013

Many of these have been corrected in the second and later printings. Errata in red are contentful and important.

• Preface
  • p. xv, ln. -21 Capital V in Frank Van Eynde
  • p. xv, ln. 3: replace 'develped' with 'developed'
• Chapter 1
  • p. 2, ln. -12: 'Will' should not be capitalized
• Chapter 2
  • p. 29, 1st line in Sec 2.4: replace 'sections' with 'section'
  • p. 42, ln. 7: replace 'the the' with 'the'
• Chapter 3
  • p. 50, ln. -19: 'interesect' -> 'intersect'
  • p. 52, ln. -7: 'deparments' -> 'departments'
  • p. 53, ln. 6: replace 'hierararchy' with 'hierarchy'
  • p. 56, fn. 7: 'devlop' -> 'develop'
  • p. 62, ln. 2: add footnote: In order to simplify our presentation, we postpone all discussion of the feature AUX until Chapter 13.
  • p. 69, ln. 14: 'so so' -> 'so'
  • p. 78, above (63): insert 'the' between 'that' and 'lexical entry'.
  • p. 79 (64b): To be fully resolved, this feature structure should also add a value for AUX ([AUX -]). However, we are postponig all discussion of AUX until Chapter 13.
  • p. 83, ln. 2: 'digrams' -> 'diagrams'
  • p. 89, Problem 3, part C: Change end to: 'any identities that the grammar (including your modified HSR2 and the Head Feature Principle) requires.'
  • p. 91 problem 5: There's a faulty presupposition in part D of this question in the version given in the first printing. In the second printing, this has been corrected dto "Does your answer to part (C) support treating five as the head of two hundred five? Briefly explain why."
  • p. 546, answer to exercise 3: In fact, phrases licensed by the rule in (47) [page 69] are incompatible with the head daughter positions of the rules in (37) [page 65] even without the specification [COMPS itr] because the former is of type phrase and the latter of type word.
• Chapter 4
  • p. 102, ln. -7 'that mother' -> 'that the mother'
  • p. 106, ln. -2 'now its final form' -> 'now in its final form'
  • p. 107 (32) The constraint on the SPR value should be HEAD.AGR [1], not just AGR [1]. While we will sometimes use such "abbreviated" pathnames, it doesn't make sense to do so in the initial presentation of a constraint such as this one.
  • p. 108 (33) Likewise, in this example, the constraint on the SPR value should be HEAD.AGR [1], not just AGR [1].
  • p. 120 There should be a definition provided for the abbreviation PP, as it is used in the Head-Modifier Rule (63). The definition should be analogous to the one given on p. 157 in Ch 5, but without the feature SYN, which is not part of the Ch 4 grammar.
  • p. 123, Problem 1, Part B: 'the data in (10d-h)' -> 'the data in (10e-h)'.
• Chapter 5
  • p. 147, below (33): The value of MOD is not an expression, but rather a list of zero or more expressions (in fact, zero or one expressions).
  • p. 147, ln. 11 Extraneous space between for and adverbs.
  • p. 151, ln -11 (46a.b) should be (46a,b).
  • p. 155, first line of 5.10.1: "in (55)" should be "below"
  • p. 161, ln. 5: replace '5.3' with '5.8'
• Chapter 6
  • p. 184, (24): The predication on the RESTR list of letters should not have [ SIT s ].
  • p. 187, (28): The value of HEAD for either the P node above to or the PP should say prep
  • p. 191, ln. -12: extraneous space before footnote mark 11.
  • p. 194, ln. 3: replace the `is-an-element-of' symbol (), with `=`.
• Chapter 7
  • p. 225, first line of part D. "(vii) and (viii) should" be "(ix) and (x)"
• Chapter 8
  • p. 238, ln. 4: Section 8.4.3 should be Section 8.4.1.
  • p. 239, (32): The DP on the ARG-ST should also have [ INDEX i ].
  • p. 242, under (39) missing "which" after "structures"
  • p. 260, (76) The type of the input should be verb-lxm to be consistent with Chapter 9, (15) on p.283.
  • p. 261, (78) should have only one feature ARG-ST, at the outermost level (not inside SYN) with the value < DP[COUNT +]_i (, PP[of]_j) >.
  • p. 261, the [RELN drive] predication on the RESTR list of (78) should have an additional feature SIT with the value s.
• Chapter 9
  • p. 290, ln. 5: replace the `is-an-element-of' symbol (), with `=`.
  • p. 301, ln. 22: interacts -> interact
  • p. 306, ln. 23: has sometimes suggested -> has sometimes been suggested
  • p. 308, first paragraph: The Argument Realization Principle does not belong under the heading of "Phrasal Licensing", and we did not mention earlier that it might vary across languages.
• Chapter 10
  • p. 313, in (8): OUPUT -> OUTPUT
  • p. 315 noteworty -> noteworthy
  • p. 328, in Passive Lexical Rule: OUPUT -> OUTPUT
• Chapter 11
  • p. 334 o with two accents should be ö in (6b) piros-fehér-zöld.
  • p. 339 "the b-sentences in (17)--(21)" should be "the b-sentences in (15)--(21)"
  • p. 345 Extraneous dot after "Semantic Inheritance Principle".
  • p. 352, in Passive Lexical Rule: OUPUT -> OUTPUT
  • p. 352, in Present Participle Lexical Rule: ARG-ST of INPUT and OUTPUT should be identified, as in the grammar summaries in Ch 9 and Appendix A.
  • p. 357 Problem 11.3, Part B: "be sure to the INDEX" -> "be sure to show the INDEX"
• Chapter 13
  • p. 408 The lexical entry in (53) is missing a type. It should be adv-lxm.
  • p. 420 The type dervv-lxm should specify [ AUX + ] so as not to inherit [ AUX - ] from verb-lxm.
  • p. 421 We should include a lexical entry for the adverb not in this grammar summay, as given in (53) on page 408. The type for that entry should be adv-lxm.
  • p. 439 The lexical entry for easy is missing its SEM value. It should have a RESTR list with one item on it, with the following constraints: [ RELN easy, SIT s_1, ARG s_2 ]. s_1 should be the value of INDEX and s_2 the value of the second argument's INDEX.
  • p. 448 The lexical entry for easy is missing its SEM value. It should have a RESTR list with one item on it, with the following constraints: [ RELN easy, SIT s_1, ARG s_2 ]. s_1 should be the value of INDEX and s_2 the value of the second argument's INDEX.
• Chapter 15
  • p. 463, fn 11: missing 1sing and 3sing before two different occurrences of `and'
  • p. 463, ln 12: `conunterpart' should be `counterpart'
  • p. 463, ln -12: `destructively modifies' should be `destructively modify'
• Chapter 16
  • p. 473, sec 16.3, ln 9: `Inuktikut' should be `Inuktitut'
• Appendices
  • p. 506, in (16): OUPUT -> OUTPUT
  • p. 517 Lex entry for not has the wrong RELN value on the predication on its RESTR list. It should be [RELN not], not [RELN today].
  • p. 517 The lexical entry for easy is missing its SEM value. It should have a RESTR list with one item on it, with the following constraints: [ RELN easy, SIT s_1, ARG s_2 ]. s_1 should be the value of INDEX and s_2 the value of the second argument's INDEX.
  • p. 520, ln. 5: replace the `is-an-element-of' symbol (), with `=`.
  • p. 531, ln. 8: replace 'a analyses' with either 'analyses' or 'an analysis'
  • p. 534, end of 2nd to last para: extraneous ,
  • p. 547, answer to Ex 2: The entry for happy should say < (PP|CP)> not S because that they have ice cream is a CP and not an S. We don't introduce complementizers until Ch 11, however. An alternative fix is to change the example in (10c) to read The children are happy they have ice cream.
  • p. 550, ln 11: SPR < [ HEAD det ] > should be ARG-ST < [ HEAD det ] >; analogously for < [ COUNT + ] >.
  • p. 561, ln 18: replace 'strcutures' with 'structures'. (Also in this entry for feature structure, we should have explained the notion of typing of feature structures.)