diff --git a/org.eclipse.xtext.xtext.generator/src/org/eclipse/xtext/xtext/generator/parser/antlr/hoisting/HoistingProcessor.java b/org.eclipse.xtext.xtext.generator/src/org/eclipse/xtext/xtext/generator/parser/antlr/hoisting/HoistingProcessor.java
index 51df56ba9..9ad351e47 100644
--- a/org.eclipse.xtext.xtext.generator/src/org/eclipse/xtext/xtext/generator/parser/antlr/hoisting/HoistingProcessor.java
+++ b/org.eclipse.xtext.xtext.generator/src/org/eclipse/xtext/xtext/generator/parser/antlr/hoisting/HoistingProcessor.java
@@ -238,7 +238,7 @@ public class HoistingProcessor {
 			// all paths are trivial
 			
 			// if there is a terminal on all branches set has terminal to true
-			// else we need might need to consider the following tokens in the path
+			// else we might need to consider the following tokens in the path
 			if (guards.stream().allMatch(HoistingGuard::hasTerminal)) {
 				return HoistingGuard.terminal();
 			} else {
diff --git a/org.eclipse.xtext.xtext.generator/src/org/eclipse/xtext/xtext/generator/parser/antlr/hoisting/pathAnalysis/TokenAnalysis.java b/org.eclipse.xtext.xtext.generator/src/org/eclipse/xtext/xtext/generator/parser/antlr/hoisting/pathAnalysis/TokenAnalysis.java
index 43a48273a..41953d2da 100644
--- a/org.eclipse.xtext.xtext.generator/src/org/eclipse/xtext/xtext/generator/parser/antlr/hoisting/pathAnalysis/TokenAnalysis.java
+++ b/org.eclipse.xtext.xtext.generator/src/org/eclipse/xtext/xtext/generator/parser/antlr/hoisting/pathAnalysis/TokenAnalysis.java
@@ -500,7 +500,7 @@ public class TokenAnalysis {
 		tokenCombinations(indexList -> {
 			log.info("current index list: " + indexList);
 			
-			// no context analysis
+			// no context analysis // TODO why?
 			List<List<Token>> tokenListsForPath = getTokenPaths(element, indexList, false);
 			List<List<Token>> tokenListForContext = getTokenPathsContextOnly(element, indexList);
 				
@@ -572,7 +572,8 @@ public class TokenAnalysis {
 	}
 	
 	public List<List<AbstractElement>> getAllPossiblePaths(AbstractElement path) {
-		return getTokenPaths(path, new TokenAnalysisPaths(range(0, config.getTokenLimit() + 1)), false, false, true)
+		// token limit + 2 so identity analysis will recognize paths that are identical up to the token limit on the flattened tree
+		return getTokenPaths(path, new TokenAnalysisPaths(range(0, config.getTokenLimit() + 2)), false, false, true)
 				.getTokenPaths()
 				.stream()
 				.map(l -> l.stream()