diff --git a/lex.go b/lex.go
index ee8c36f..3f47c2e 100644
--- a/lex.go
+++ b/lex.go
@@ -27,19 +27,28 @@ func (t tokenType) String() string {
 		return "t_equals"
 	case t_comment:
 		return "t_comment"
+	case t_list_start:
+		return "t_list_start"
+	case t_list_end:
+		return "t_list_end"
+	case t_list_separator:
+		return "t_list_separator"
 	default:
 		panic(fmt.Sprintf("unknown token type: %v", t))
 	}
 }
 
 const (
-	t_error   tokenType = iota // a stored lex error
-	t_eof                      // end of file token
-	t_string                   // a string literal
-	t_name                     // a name
-	t_type                     // a type
-	t_equals                   // equals sign
-	t_comment                  // a comment
+	t_error          tokenType = iota // a stored lex error
+	t_eof                             // end of file token
+	t_string                          // a string literal
+	t_name                            // a name
+	t_type                            // a type
+	t_equals                          // equals sign
+	t_comment                         // a comment
+	t_list_start                      // [
+	t_list_end                        // ]
+	t_list_separator                  // ,
 )
 
 type stateFn func(*lexer) (stateFn, error)
@@ -140,6 +149,18 @@ func lexRoot(l *lexer) (stateFn, error) {
 		return lexStringLiteral(r), nil
 	case r == '#':
 		return lexComment, nil
+	case r == '[':
+		l.keep(r)
+		l.emit(t_list_start)
+		return lexRoot, nil
+	case r == ']':
+		l.keep(r)
+		l.emit(t_list_end)
+		return lexRoot, nil
+	case r == ',':
+		l.keep(r)
+		l.emit(t_list_separator)
+		return lexRoot, nil
 	case unicode.IsSpace(r):
 		return lexRoot, nil
 	case unicode.IsLower(r):
diff --git a/lex_test.go b/lex_test.go
index 60e8baf..db669a5 100644
--- a/lex_test.go
+++ b/lex_test.go
@@ -35,6 +35,8 @@ var primitivesTests = []struct {
     # comment line one
     # comment line two
     `, []token{{t_comment, " comment line one"}, {t_comment, " comment line two"}}},
+	{`[]`, []token{{t_list_start, "["}, {t_list_end, "]"}}},
+	{`["item"]`, []token{{t_list_start, "["}, {t_string, "item"}, {t_list_end, "]"}}},
 }
 
 func TestLexPrimities(t *testing.T) {
diff --git a/node.go b/node.go
index 399dd4b..cf3b4a1 100644
--- a/node.go
+++ b/node.go
@@ -67,7 +67,7 @@ func (n *rootNode) String() string {
 	for _, child := range n.children {
 		fmt.Fprintf(&buf, "%s, ", child)
 	}
-	if len(n.children) > 0 {
+	if buf.Len() > 1 {
 		buf.Truncate(buf.Len() - 2)
 	}
 	buf.WriteString("}")
@@ -109,20 +109,51 @@ func (n *assignmentNode) parse(p *parser) error {
 		return fmt.Errorf("parse error: unexpected %v token after name, expected =", t.t)
 	}
 
-	t = p.next()
-	switch t.t {
-	case t_error:
-		return fmt.Errorf("parse error: saw lex error while parsing assignment node: %v", t.s)
-	case t_eof:
-		return fmt.Errorf("parse error: unexpected eof in assignment node")
-	case t_string:
-		n.value = t.s
-		return nil
-	default:
-		return fmt.Errorf("parse error: unexpected %v token after =, expected some kind of value", t.t)
+	v, err := p.parseValue()
+	if err != nil {
+		return err
 	}
+	n.value = v
+	return nil
 }
 
 func (n *assignmentNode) String() string {
 	return fmt.Sprintf("{assign: name=%s, val=%s}", n.name, n.value)
 }
+
+type list struct {
+	head *listElem
+	tail *listElem
+}
+
+func (l list) String() string {
+	var buf bytes.Buffer
+	buf.WriteString("[")
+	for e := l.head; e != nil; e = e.next {
+		fmt.Fprintf(&buf, "%v, ", e.value)
+	}
+	if buf.Len() > 1 {
+		buf.Truncate(buf.Len() - 2)
+	}
+	buf.WriteString("]")
+	return buf.String()
+}
+
+func (l *list) append(v interface{}) {
+	e := listElem{value: v}
+	if l.head == nil {
+		l.head = &e
+	}
+
+	if l.tail != nil {
+		l.tail.next = &e
+		e.prev = l.tail
+	}
+	l.tail = &e
+}
+
+type listElem struct {
+	value interface{}
+	prev  *listElem
+	next  *listElem
+}
diff --git a/parse.go b/parse.go
index f6178c8..56fa44a 100644
--- a/parse.go
+++ b/parse.go
@@ -1,6 +1,7 @@
 package main
 
 import (
+	"fmt"
 	"io"
 )
 
@@ -40,3 +41,57 @@ func (p *parser) next() token {
 	}
 	return <-p.input
 }
+
+func (p *parser) peek() token {
+	t := p.next()
+	p.unread(t)
+	return t
+}
+
+func (p *parser) unread(t token) {
+	if p.backup == nil {
+		p.backup = make([]token, 0, 8)
+	}
+	p.backup = append(p.backup, t)
+}
+
+// parse the next value.  This is to be executed in a context where we know we
+// want something that is a value to come next, such as after an equals sign.
+func (p *parser) parseValue() (interface{}, error) {
+	for {
+		t := p.next()
+		switch t.t {
+		case t_error:
+			return nil, fmt.Errorf("parse error: saw lex error when looking for value: %v", t.s)
+		case t_eof:
+			return nil, fmt.Errorf("parse error: unexpected eof when looking for value")
+		case t_string:
+			return t.s, nil
+		case t_list_start:
+			l := new(list)
+
+		SIN:
+			if p.peek().t == t_list_end {
+				p.next()
+				return l, nil
+			}
+
+			if v, err := p.parseValue(); err != nil {
+				return nil, err
+			} else {
+				l.append(v)
+			}
+
+			switch t := p.next(); t.t {
+			case t_list_separator:
+				goto SIN
+			case t_list_end:
+				return l, nil
+			default:
+				return nil, fmt.Errorf("parse error: unexpected %v token while scanning for list", t.t)
+			}
+		default:
+			return nil, fmt.Errorf("parse error: unexpected %v token while looking for value", t.t)
+		}
+	}
+}
diff --git a/parse_test.go b/parse_test.go
index 544ef09..ef123e8 100644
--- a/parse_test.go
+++ b/parse_test.go
@@ -53,6 +53,12 @@ var parseTests = []parseTest{
         `,
 		root: &rootNode{},
 	},
+	{
+		source: `
+        heroes = ["lina", "cm"]
+        `,
+		root: &rootNode{},
+	},
 }
 
 type parseTest struct {