pronounced as /notice/Bleeding order is a term used in phonology to describe specific interactions of phonological rules. The term was introduced in 1968 by Paul Kiparsky.[1] If two phonological rules are said to be in bleeding order, the application of the first rule creates a context in which the second rule can no longer apply.
The opposite of this is called feeding order.
An example of this in English is the pronounced as //ɪ//-insertion between a voiceless alveolar fricative and a plural-z, as in buses pronounced as /[bʌsɪz]/ (with the underlying representation pronounced as ///bʌs-z///). English also has a rule which devoices segments after voiceless consonants, as in books pronounced as /[bʊks]/, with the underlying representation pronounced as ///bʊk-z///). In the output form pronounced as /[bʌsɪz]/ buses, final devoicing has not applied, because the phonological context in which this rule could have applied has by the application the application of pronounced as //ɪ//-insertion. Put differently, the application order "(1) pronounced as //ɪ//-insertion (2) final devoicing" is a bleeding order in English.
If two rules which would have a bleeding relationship in one order actually apply in the opposite order, the latter is called a counterbleeding order. An example of this can be seen in the pronunciation of the diminutive of the word slang ("snake") in the Dutch dialect of Kaatsheuvel: pronounced as /[slɑŋəskə]/. If [s]-insertion had applied first, then the rule which inserts an additional pronounced as //-ə// between the noun stem and the suffix pronounced as //-kə// could no longer have applied and the output form would have been pronounced as /[slɑŋskə]/. However, the rules have applied in the reverse order.
Amsterdam: J. Benjamins ISBN 978-90-272-7517-2.