Editing Old Japanese (section)

===Morphophonemics===
Two adjacent vowels fused to form a new vowel when a consonant was lost within a morpheme, or a compound was lexicalized as a single morpheme.
The following fusions occurred:

;''i<sub>1</sub>'' + ''a'' → ''e<sub>1</sub>''
:* {{tlit|ojp|saki<sub>1</sub>}} 'bloom' + {{tlit|ojp|ari}} 'exist' → {{tlit|ojp|sake<sub>1</sub>ri}} 'be blooming'{{sfn|Russell|2003|p=532}}{{sfn|Frellesvig|2010|p=48}}
:* {{tlit|ojp|ki<sub>1</sub>}} 'wear' + {{tlit|ojp|aru}} 'be.{{sc|ATT}}' → {{tlit|ojp|ke<sub>1</sub>ru}} 'wear.{{sc|FIN}}'{{sfn|Miyake|2003|p=81}}
:Further examples are provided by verbs ending with the retrospective auxiliary -{{tlit|ojp|ki<sub>1</sub>}} and the verbal suffixes {{tlit|ojp|amu}} 'conjecture' or {{tlit|ojp|ari}} 'exist':{{sfn|Russell|2003|pp=521–522}}
:* {{tlit|ojp|tir-i-ki<sub>1</sub> 'fall, scatter.{{sc|INF}}.{{sc|RET}}'}} + {{tlit|ojp|am-u}} → {{tlit|ojp|tirike<sub>1</sub>mu}} '(it) has surely fallen'{{sfn|Russell|2003|p=531}}
:* {{tlit|ojp|ari-ki<sub>1</sub>}} 'exist.{{sc|INF}}.{{sc|RET}}' + {{tlit|ojp|ar-i}} → {{tlit|ojp|arike<sub>1</sub>ri}} 'it existed'{{sfn|Russell|2003|p=531}}
;''i<sub>1</sub>'' + ''o<sub>2</sub>'' → ''e<sub>1</sub>''
:* {{tlit|ojp|utusi}} 'real' + {{tlit|ojp|oyomi<sub>1</sub>}} 'person' → {{tlit|ojp|utuse<sub>1</sub>mi<sub>1</sub>}} 'living person'{{sfn|Erickson|2003|p=499}}{{efn|An alternative form, {{tlit|ojp|utuso<sub>2</sub>mi<sub>1</sub>}}, obtained by deleting the vowel ''-i'', is also attested.{{sfn|Erickson|2003|p=499}}}}
;''a'' + ''i'' → ''e<sub>2</sub>''
:* {{tlit|ojp|naga}} 'long' + {{tlit|ojp|iki<sub>1</sub>}} 'breath' → {{tlit|ojp|nage<sub>2</sub>ki<sub>1</sub>}} 'sigh'{{sfn|Russell|2003|p=532}}
:* {{tlit|ojp|taka}} 'high' + {{tlit|ojp|iti}} 'market' → {{tlit|ojp|take<sub>2</sub>ti}} (place name){{sfn|Russell|2003|p=532}}{{sfn|Frellesvig|2010|p=46}}
;''o<sub>2</sub>'' + ''i'' → ''e<sub>2</sub>''
:* {{tlit|ojp|tono}} 'palace' + {{tlit|ojp|iri}} 'enter' → {{tlit|ojp|toneri}} 'attendant'{{sfn|Russell|2003|p=532}}{{sfn|Miyake|2003|p=81}}
;''o<sub>2</sub>'' + ''i'' → ''i<sub>2</sub>''
:* {{tlit|ojp|opo}} 'big' + {{tlit|ojp|isi}} 'rock' → {{tlit|ojp|opi<sub>2</sub>si}} 'big rock'{{sfn|Russell|2003|p=532}}{{sfn|Miyake|2003|p=81}}
;''u'' + ''i'' → ''i<sub>2</sub>''
:* {{tlit|ojp|waku}} 'young' + {{tlit|ojp|iratuko<sub>1</sub>}} 'term of veneration (male)' → {{tlit|ojp|waki<sub>2</sub>ratuko<sub>1</sub>}} (title){{sfn|Russell|2003|p=532}}{{sfn|Miyake|2003|p=81}}
;''u'' + ''a'' → ''o<sub>1</sub>''
:* {{tlit|ojp|kazu}} 'number' + {{tlit|ojp|ape<sub>2</sub>}} 'to join' → {{tlit|ojp|kazo<sub>1</sub>pe<sub>2</sub>}} 'to count'{{sfn|Frellesvig|2010|p=48}}{{sfn|Russell|2003|p=531}}
;''u'' + ''o'' → ''o<sub>1</sub>''
:* {{tlit|ojp|situ}} 'ancient type of native weaving' + {{tlit|ojp|ori}} 'weaving' → {{tlit|ojp|sito<sub>1</sub>ri}} 'native weaving'{{sfn|Russell|2003|p=532}}{{sfn|Frellesvig|2010|p=48}}

Adjacent vowels belonging to different morphemes, or pairs of vowels for which none of the above fusions applied, were reduced by deleting one or other of the vowels.{{sfn|Russell|2003|p=523}}
Most often, the first of the adjacent vowels was deleted:{{sfn|Unger|2000|p=662}}{{sfn|Vovin|2020|p=55}}
* {{tlit|ojp|to<sub>2</sub>ko<sub>2</sub>}} 'eternal' + {{tlit|ojp|ipa}} 'rock' → {{tlit|ojp|to<sub>2</sub>ki<sub>1</sub>pa}} 'eternal rock; everlasting'{{sfn|Russell|2003|p=528}}{{sfn|Frellesvig|2010|p=40}}
* {{tlit|ojp|ama}} 'heaven' + {{tlit|ojp|ori}} 'descend' → {{tlit|ojp|amori}} 'descend from heaven'{{sfn|Russell|2003|p=525}}
The exception to this rule occurred when the first of the adjacent vowels was the sole vowel of a monosyllabic morpheme (usually a [[clitic]]), in which case the other vowel was deleted:{{sfn|Unger|2000|p=662}}{{sfn|Vovin|2020|p=55}}
* {{tlit|ojp|mi<sub>1</sub>}} (honorific) + {{tlit|ojp|uma}} 'horse' → {{tlit|ojp|mi<sub>1</sub>ma}} 'honourable horse'{{sfn|Russell|2003|p=530}}
* {{tlit|ojp|ko<sub>1</sub>}} 'child, egg' + {{tlit|ojp|umu}} 'birth' → {{tlit|ojp|ko<sub>1</sub>mu}} 'give birth, lay an egg'{{sfn|Russell|2003|p=529}}
Cases where both outcomes are found are attributed to different analyses of morpheme boundaries:{{sfn|Unger|2000|p=662}}{{sfn|Frellesvig|2010|p=40}}
* {{tlit|ojp|waga}} 'my' + {{tlit|ojp|ipe<sub>1</sub>}} 'house' → {{tlit|ojp|wagi<sub>1</sub>pe<sub>1</sub>}} 'my house'
* {{tlit|ojp|wa}} 'I' + {{tlit|ojp|ga}} {{sc|GEN}} + {{tlit|ojp|ipe<sub>1</sub>}} 'house' → {{tlit|ojp|wagape<sub>1</sub>}} 'my house'