My number is six thirty eight.

This phenomenon concerns sequences of numbers that are pronounced not as one single (large) number, but as a string of smaller ones. This is typical in the spoken rendering of digits used for telephone numbers, postal codes, and other identifiers. For example, given the zip code 94305, most American English speakers would read that aloud as nine four three oh five or nine four three zero five, and not ninety-four thousand three hundred and five.

The following are examples involving this phenomenon:

• Nine six seven five four oh nine.

• The zip code is nine four three zero five.

On the other hand, sequences of digits read as representing one number are not examples of the number sequence phenomenon. Such examples are discussed under the heading `Compositional Number Names.'

• The crate contained four thousand five hundred and seven rubber duckies.

Analyses of strings involving this phenomenon are characterized by the EP num_seq.

  num_seq[L-INDEX x1, R-INDEX x2]

The variables x1 and x2 above are shared either with the ARG0 of card predicates or the ARG0 of another num_seq; like with coordination, number sequences with three or more constituent numbers are treated via successive binary combination.

The numbers in number sequences can themselves be compositional number names, as in the testsuite example above or the following:

• The treaty was signed in nineteen eight-nine.

• The example of year names shows that the choice to read a sequence of digits as one larger number or a sequence of smaller onces does not correlate with whether the item identified by the sequence is saliently part of a series numbered in order.

• Change the testsuite example? six thirty eight is ambiguous between a reading with one instance of num_seq and one with two.

• Change the argument names in num_seq, from L-INDEX and R-INDEX to ARG1 and ARG2?

• I changed the name of the phenomenon from Number Expression to Number Sequences, since the former seemed too general.

• 1212

• ErgSemantics main page

• Inventory of semantic phenomena (to be) documented

• How to cite this work