diff --git a/extra/io/encodings/iana/iana.factor b/extra/io/encodings/iana/iana.factor
old mode 100644
new mode 100755
index 24badaf683..dd429c1670
--- a/extra/io/encodings/iana/iana.factor
+++ b/extra/io/encodings/iana/iana.factor
@@ -1,8 +1,8 @@
 ! Copyright (C) 2008 Daniel Ehrenberg
 ! See http://factorcode.org/license.txt for BSD license.
-USING: kernel strings unicode.syntax.backend io.files assocs
-splitting sequences io namespaces sets
-io.encodings.ascii io.encodings.utf8 io.encodings.utf16 io.encodings.8-bit ;
+USING: kernel strings values io.files assocs
+splitting sequences io namespaces sets io.encodings.8-bit
+io.encodings.ascii io.encodings.utf8 io.encodings.utf16  ;
 IN: io.encodings.iana
 
 <PRIVATE
diff --git a/extra/unicode/breaks/breaks.factor b/extra/unicode/breaks/breaks.factor
index d8e4f8c24e..3787f78648 100755
--- a/extra/unicode/breaks/breaks.factor
+++ b/extra/unicode/breaks/breaks.factor
@@ -1,7 +1,7 @@
 USING: unicode.categories kernel math combinators splitting
 sequences math.parser io.files io assocs arrays namespaces
-math.ranges unicode.normalize unicode.syntax.backend
-unicode.syntax unicode.data compiler.units alien.syntax io.encodings.ascii ;
+math.ranges unicode.normalize values io.encodings.ascii
+unicode.syntax unicode.data compiler.units alien.syntax  ;
 IN: unicode.breaks
 
 C-ENUM: Any L V T Extend Control CR LF graphemes ;
diff --git a/extra/unicode/data/data.factor b/extra/unicode/data/data.factor
index 85ce50acb9..d630aacbed 100755
--- a/extra/unicode/data/data.factor
+++ b/extra/unicode/data/data.factor
@@ -1,7 +1,7 @@
 USING: assocs math kernel sequences io.files hashtables
 quotations splitting arrays math.parser hash2 math.order
 byte-arrays words namespaces words compiler.units parser
-io.encodings.ascii unicode.syntax.backend ;
+io.encodings.ascii values ;
 IN: unicode.data
 
 ! Convenience functions
diff --git a/extra/unicode/script/script.factor b/extra/unicode/script/script.factor
index d0bb4ac30d..846f797f71 100755
--- a/extra/unicode/script/script.factor
+++ b/extra/unicode/script/script.factor
@@ -1,4 +1,4 @@
-USING: unicode.syntax.backend kernel sequences assocs io.files
+USING: values kernel sequences assocs io.files
 io.encodings ascii math.ranges io splitting math.parser 
 namespaces byte-arrays locals math sets io.encodings.ascii
 words compiler.units arrays interval-maps ;
diff --git a/extra/values/authors.txt b/extra/values/authors.txt
new file mode 100755
index 0000000000..504363d316
--- /dev/null
+++ b/extra/values/authors.txt
@@ -0,0 +1 @@
+Daniel Ehrenberg
diff --git a/extra/values/summary.txt b/extra/values/summary.txt
new file mode 100755
index 0000000000..7caab7412d
--- /dev/null
+++ b/extra/values/summary.txt
@@ -0,0 +1 @@
+Global variables in the Forth value style
diff --git a/extra/values/tags.txt b/extra/values/tags.txt
new file mode 100755
index 0000000000..187b6926c1
--- /dev/null
+++ b/extra/values/tags.txt
@@ -0,0 +1 @@
+extensions
diff --git a/extra/values/values-docs.factor b/extra/values/values-docs.factor
new file mode 100755
index 0000000000..4984b03f03
--- /dev/null
+++ b/extra/values/values-docs.factor
@@ -0,0 +1,27 @@
+USING: help.markup help.syntax ;
+IN: values
+
+ARTICLE: "values" "Global values"
+"Usually, dynamically scoped variables are sufficient for holding data which is not literal. But occasionally, for global information that's calculated just once, it's useful to use the word mechanism instead, and set the word to the appropriate value just once. Values abstract over this concept. To create a new word as a value, use the following syntax:"
+{ $subsection POSTPONE: VALUE: }
+"To get the value, just call the word. The following words manipulate values:"
+{ $subsection get-value }
+{ $subsection set-value }
+{ $subsection change-value } ;
+
+HELP: VALUE:
+{ $syntax "VALUE: word" }
+{ $values { "word" "a word to be created" } }
+{ $description "Creates a value on the given word, initializing it to hold " { $code f } ". To get the value, just run the word. To set it, use " { $link set-value } "." } ;
+
+HELP: get-value
+{ $values { "word" "a value word" } { "value" "the contents" } }
+{ $description "Gets a value. This should not normally be used, unless the word is not known until runtime." } ;
+
+HELP: set-value
+{ $values { "value" "a new value" } { "word" "a value word" } }
+{ $description "Sets the value word." } ;
+
+HELP: change-value
+{ $values { "word" "a value word" } { "quot" "a quotation ( oldvalue -- newvalue )" } }
+{ $description "Changes the value using the given quotation." } ;
diff --git a/extra/values/values-tests.factor b/extra/values/values-tests.factor
new file mode 100755
index 0000000000..31b44be99e
--- /dev/null
+++ b/extra/values/values-tests.factor
@@ -0,0 +1,9 @@
+USING: tools.test values math ;
+IN: values.tests
+
+VALUE: foo
+[ f ] [ foo ] unit-test
+[ ] [ 3 \ foo set-value ] unit-test
+[ 3 ] [ foo ] unit-test
+[ ] [ \ foo [ 1+ ] change-value ] unit-test
+[ 4 ] [ foo ] unit-test
diff --git a/extra/unicode/syntax/backend/backend.factor b/extra/values/values.factor
old mode 100644
new mode 100755
similarity index 52%
rename from extra/unicode/syntax/backend/backend.factor
rename to extra/values/values.factor
index 5c463e8fc4..0d1ea3bc04
--- a/extra/unicode/syntax/backend/backend.factor
+++ b/extra/values/values.factor
@@ -1,8 +1,14 @@
 USING: kernel parser sequences words ;
-IN: unicode.syntax.backend
+IN: values
 
 : VALUE:
     CREATE-WORD { f } clone [ first ] curry define ; parsing
 
 : set-value ( value word -- )
     word-def first set-first ;
+
+: get-value ( word -- value )
+    word-def first first ;
+
+: change-value ( word quot -- )
+    over >r >r get-value r> call r> set-value ; inline