diff --git a/edb/common/parsing.py b/edb/common/parsing.py
index 1ec8b2363e6..784830c1087 100644
--- a/edb/common/parsing.py
+++ b/edb/common/parsing.py
@@ -25,7 +25,6 @@
 import json
 import logging
 import os
-import sys
 import types
 
 import parsing
@@ -46,13 +45,9 @@ class Token(parsing.Token):
     _token: str = ""
 
     def __init_subclass__(
-            cls, *, token=None, lextoken=None, precedence_class=None,
-            is_internal=False, **kwargs):
+            cls, *, token=None, lextoken=None, is_internal=False, **kwargs):
         super().__init_subclass__(**kwargs)
 
-        if precedence_class is not None:
-            cls._precedence_class = precedence_class
-
         if is_internal:
             return
 
@@ -72,18 +67,7 @@ def __init_subclass__(
         if not cls.__doc__:
             doc = '%%token %s' % token
 
-            pcls = getattr(cls, '_precedence_class', None)
-            if pcls is None:
-                try:
-                    pcls = sys.modules[Token.__module__].PrecedenceMeta
-                except (KeyError, AttributeError):
-                    pass
-
-            if pcls is None:
-                msg = 'Precedence class is not set for {!r}'.format(Token)
-                raise TypeError(msg)
-
-            prec = pcls.for_token(token)
+            prec = Precedence.for_token(token)
             if prec:
                 doc += ' [%s]' % prec.__name__
 
@@ -235,55 +219,44 @@ def decorator(func):
     return decorator
 
 
-class PrecedenceMeta(type):
-    token_prec_map: Dict[Tuple[Any, Any], Any] = {}
-    last: Dict[Tuple[Any, Any], Any] = {}
+class Precedence(parsing.Precedence):
+    token_prec_map: Dict[Any, Any] = {}
+    last: Dict[Any, Any] = {}
 
-    def __new__(
-            mcls, name, bases, dct, *, assoc, tokens=None, prec_group=None,
-            rel_to_last='>'):
-        result = super().__new__(mcls, name, bases, dct)
+    def __init_subclass__(
+            cls, *, assoc, tokens=None, prec_group=None, rel_to_last='>',
+            is_internal=False, **kwargs):
+        super().__init_subclass__(**kwargs)
 
-        if name == 'Precedence':
-            return result
+        if is_internal:
+            return
 
-        if not result.__doc__:
+        if not cls.__doc__:
             doc = '%%%s' % assoc
 
-            last = mcls.last.get((mcls, prec_group))
+            last = Precedence.last.get(prec_group)
             if last:
                 doc += ' %s%s' % (rel_to_last, last.__name__)
 
-            result.__doc__ = doc
+            cls.__doc__ = doc
 
         if tokens:
             for token in tokens:
                 existing = None
                 try:
-                    existing = mcls.token_prec_map[mcls, token]
+                    existing = Precedence.token_prec_map[token]
                 except KeyError:
-                    mcls.token_prec_map[mcls, token] = result
+                    Precedence.token_prec_map[token] = cls
                 else:
                     raise Exception(
                         'token {} has already been set precedence {}'.format(
                             token, existing))
 
-        mcls.last[mcls, prec_group] = result
-
-        return result
-
-    def __init__(
-            cls, name, bases, dct, *, assoc, tokens=None, prec_group=None,
-            rel_to_last='>'):
-        super().__init__(name, bases, dct)
+        Precedence.last[prec_group] = cls
 
     @classmethod
-    def for_token(mcls, token_name):
-        return mcls.token_prec_map.get((mcls, token_name))
-
-
-class Precedence(parsing.Precedence, assoc='fail', metaclass=PrecedenceMeta):
-    pass
+    def for_token(cls, token_name):
+        return Precedence.token_prec_map.get(token_name)
 
 
 def load_parser_spec(
diff --git a/edb/edgeql/parser/grammar/precedence.py b/edb/edgeql/parser/grammar/precedence.py
index f324f68003d..414cf8dccad 100644
--- a/edb/edgeql/parser/grammar/precedence.py
+++ b/edb/edgeql/parser/grammar/precedence.py
@@ -22,11 +22,7 @@
 from edb.common import parsing
 
 
-class PrecedenceMeta(parsing.PrecedenceMeta):
-    pass
-
-
-class Precedence(parsing.Precedence, assoc='fail', metaclass=PrecedenceMeta):
+class Precedence(parsing.Precedence, assoc='fail', is_internal=True):
     pass
 
 
diff --git a/edb/edgeql/parser/grammar/tokens.py b/edb/edgeql/parser/grammar/tokens.py
index 40110efcc67..0c4b56d6245 100644
--- a/edb/edgeql/parser/grammar/tokens.py
+++ b/edb/edgeql/parser/grammar/tokens.py
@@ -27,15 +27,13 @@
 from edb.common import parsing
 
 from . import keywords
-from . import precedence
 
 
 clean_string = re.compile(r"'(?:\s|\n)+'")
 string_quote = re.compile(r'\$(?:[A-Za-z_][A-Za-z_0-9]*)?\$')
 
 
-class Token(parsing.Token, precedence_class=precedence.PrecedenceMeta,
-            is_internal=True):
+class Token(parsing.Token, is_internal=True):
     pass