Universal #1379

Legacy number: 1379

 Number  1375
 Original  Subordinate Gap/No Gap Hierarchy:
If a matrix filler can be matched with a gap in a subordinate clause of complexity n containing another gap, then it can be matched with a gap in a subordinate clause of complexity n containing no other gap.
 Standardized  Subordinate Gap/No Gap Hierarchy:
IF a matrix filler can be matched with a gap in a subordinate clause of complexity n containing another gap, THEN it can be matched with a gap in a subordinate clause of complexity n containing no other gap.
 Formula  matrix filler & gap (subordinate clause of complexity n containing gap) ⇒ matrix filler & gap (subordinate clause of complexity n containing no gap)
 Keywords  filler, gap, complexity
 Domain  syntax
 Type  implication
 Status  achronic
 Quality  absolute
 Basis  Russian (Slavic), English, German, Swedish, Norwegian (all Germanic)
 Source  Hawkins 1999: 269

← Back to Universals Archive

Discussion
Notes
1. The terms “filler” and “gap” are used for the moved element and its trace respectively.2. Hawkins claims that this hierarchy can be explained by the following processing preference: Reduce Additional Syntactic Processing in FGDs: The human processor prefers to minimize the syntactic rules and processing operations that apply in FGDs.3. One manifestation of this preference can be seen in the WH-island effects of Chomsky 1973. A subordinate clause that has undergone the syntactic process of WH-fronting and coindexation with a gap, e.g. in an indirect question, is more difficult to extract out of than a simple that-clause in English.(i) a. What(i) did you hope [that they would bake O(i)]? b. *What(i) did you wonder [how(j) they would bake O(i) O(j)]?
Comment
Frans Plank · 3 Aug 2006
1. The terms 'filler' and 'gap' are used for the moved element and its trace respectively.

2. Hawkins claims that this hierarchy can be explained by the following processing preference: Reduce Additional Syntactic Processing in FGDs: The human processor prefers to minimize the syntactic rules and processing operations that apply in FGDs.

3. One manifestation of this preference can be seen in the WH-island effects of Chomsky 1973. A subordinate clause that has undergone the syntactic process of WH-fronting and coindexation with a gap, e.g. in an indirect question, is more difficult to extract out of than a simple that-clause in English.

<table class = "textTable">
<tr>
<td> (i)</td>
<td>a.</td>
<td> What(i) did you hope [that they would bake O(i)]?</td>
</tr>
<tr>
<td> </td>
<td>b.</td>
<td> *What(i) did you wonder [how(j) they would bake O(i) O(j)]?</td>
</tr>
</table>