diff --git a/spacy/ja/__init__.py b/spacy/ja/__init__.py
index b2ec281f7..26e39a593 100644
--- a/spacy/ja/__init__.py
+++ b/spacy/ja/__init__.py
@@ -16,12 +16,7 @@ from collections import namedtuple
 
 ShortUnitWord = namedtuple('ShortUnitWord', ['surface', 'base_form', 'part_of_speech'])
 
-class JapaneseDoc(Doc):
-    def __init__(self, detailed_tokens, vocab, words=None, spaces=None, orths_and_spaces=None):
-        super(JapaneseDoc, self).__init__(vocab, words, spaces, orths_and_spaces)
-        # This saves tokenizer output so mecab doesn't have to be called again
-        # when determining POS tags.
-        self.detailed_tokens = detailed_tokens
+DETAILS_KEY = 'mecab_details'
 
 def try_mecab_import():
     """Mecab is required for Japanese support, so check for it.
@@ -43,7 +38,10 @@ class JapaneseTokenizer(object):
     def __call__(self, text):
         dtokens = detailed_tokens(self.tokenizer, text)
         words = [x.surface for x in dtokens]
-        return JapaneseDoc(dtokens, self.vocab, words=words, spaces=[False]*len(words))
+        doc = Doc(self.vocab, words=words, spaces=[False]*len(words))
+        # stash details tokens for tagger to use
+        doc.user_data[DETAILS_KEY] = dtokens
+        return doc
 
 def resolve_pos(token):
     """If necessary, add a field to the POS tag for UD mapping.
@@ -99,7 +97,7 @@ class JapaneseTagger(object):
         # 1. get raw JP tags
         # 2. add features to tags as necessary for UD
 
-        dtokens = tokens.detailed_tokens
+        dtokens = tokens.user_data[DETAILS_KEY]
         rawtags = list(map(resolve_pos, dtokens))
         self.tagger.tag_from_strings(tokens, rawtags)