diff --git a/NLP/Tokenize.hs b/NLP/Tokenize.hs
--- a/NLP/Tokenize.hs
+++ b/NLP/Tokenize.hs
@@ -9,12 +9,14 @@
     , punctuation
     , finalPunctuation
     , initialPunctuation
+    , contractions
     , negatives
     )
 where
 
 import qualified Data.Char as Char
 import Data.List
+import Data.Maybe
 import Control.Monad.Instances
 import Data.List.Split
 import Control.Monad
@@ -39,7 +41,11 @@
 run f = map unwrap . unE . f
 
 defaultTokenizer :: Tokenizer
-defaultTokenizer = whitespace >=> uris >=> punctuation >=> negatives
+defaultTokenizer =     whitespace 
+                   >=> uris 
+                   >=> punctuation 
+                   >=> contractions 
+                   >=> negatives 
 
 -- | Detect common uris and freeze them
 uris :: Tokenizer
@@ -53,25 +59,40 @@
 
 -- | Split off word-final punctuation
 finalPunctuation :: Tokenizer
-finalPunctuation x = E $
+finalPunctuation x = E . filter (not . null . unwrap) $
     case span Char.isPunctuation . reverse $ x of
       ([],w) -> [Right . reverse $ w]
       (ps,w) -> [Right . reverse $ w, Right . reverse $ ps]
 
 -- | Split off word-initial punctuation
 initialPunctuation :: Tokenizer
-initialPunctuation x = E $
+initialPunctuation x = E . filter (not . null . unwrap) $
     case span Char.isPunctuation$ x of
       ([],w) -> [Right w]
       (ps,w) -> [Right ps, Right w]
 
--- | Split words ending in "n't", and freeze "n't" 
+-- | Split words ending in n't, and freeze n't 
 negatives :: Tokenizer
 negatives x | "n't" `isSuffixOf` x = E [ Right . reverse . drop 3 . reverse $ x
                                        , Left "n't" ]
             | True                 = E [Right x]
 
--- | Split string on whitespace. This is just a wrapper for 'words'
+-- | Split common contractions off and freeze them.
+-- | Currently deals with: 's, 'd, 've, 'll
+contractions :: Tokenizer
+contractions x = case catMaybes . map (splitSuffix x) $ cts of
+                   [] -> return x
+                   ((w,s):_) -> E [ Right w,Left s]
+    where cts = ["'s","'d","'ve","'ll"]
+          splitSuffix w sfx = 
+              let w' = reverse w
+                  len = length sfx
+              in if sfx `isSuffixOf` w 
+                 then Just (take (length w - len) w, reverse . take len $ w')
+                 else Nothing
+
+
+-- | Split string on whitespace. This is just a wrapper for Data.List.words
 whitespace :: Tokenizer
 whitespace xs = E [Right w | w <- words xs ]
 
@@ -84,7 +105,11 @@
 
 examples = 
     ["This shouldn't happen."
-    ,"Some 'quoted' struff"
+    ,"Some 'quoted' stuff"
     ,"This is a URL: http://example.org."
-    ,"How about an email@example.com"]
+    ,"How about an email@example.com"
+    ,"ReferenceError #1065 broke my debugger!"
+    ,"I would've gone."
+    ,"They've been there."
+    ]
 
diff --git a/tokenize.cabal b/tokenize.cabal
--- a/tokenize.cabal
+++ b/tokenize.cabal
@@ -7,7 +7,7 @@
 -- The package version. See the Haskell package versioning policy
 -- (http://www.haskell.org/haskellwiki/Package_versioning_policy) for
 -- standards guiding when and how versions should be incremented.
-Version:             0.1.1
+Version:             0.1.2
 
 -- A short (one-line) description of the package.
 Synopsis:            Simple tokenizer for English text.