diff --git a/23-passive-aggressive/README.md b/23-passive-aggressive/README.md
new file mode 100644
index 0000000..d416a3b
--- /dev/null
+++ b/23-passive-aggressive/README.md
@@ -0,0 +1,16 @@
+Style #23
+==============================
+
+Constraints:
+
+- Computation should not proceed if exceptions occur
+
+- Core program functions have no exception handling, not even the main function
+
+- The final result should show the exception, if it occurred
+
+Possible names:
+
+- Passive aggressive
+- Monadic Exceptions
+
diff --git a/23-passive-aggressive/tf-23.py b/23-passive-aggressive/tf-23.py
new file mode 100644
index 0000000..7d81f71
--- /dev/null
+++ b/23-passive-aggressive/tf-23.py
@@ -0,0 +1,71 @@
+#!/usr/bin/env python
+import sys, re, operator, string
+
+#
+# The PassiveAggressive class for this example
+#
+class TFPassiveAggressive:
+    def __init__(self, v):
+        self._e = None
+        self._offending_func = None
+        self._value = v
+
+    def bind(self, func):
+        if self._e == None:
+            try:
+                self._value = func(self._value)
+            except Exception as e:
+                self._e = e
+                self._offending_func = func
+        return self
+
+    def printme(self):
+        if self._e == None:
+            print self._value
+        else:
+            print self._e, " in ", self._offending_func.__name__
+
+#
+# The functions
+#
+def get_input(arg):
+    return sys.argv[1]
+
+def extract_words(path_to_file):
+    with open(path_to_file) as f:
+        data = f.read()
+    pattern = re.compile('[\W_]+')
+    word_list = pattern.sub(' ', data).lower().split()
+    return word_list
+
+def remove_stop_words(word_list):
+    with open('../stop_words.txt') as f:
+        stop_words = f.read().split(',')
+    # add single-letter words
+    stop_words.extend(list(string.ascii_lowercase))
+    return [w for w in word_list if not w in stop_words]
+
+def frequencies(word_list):
+    word_freqs = {}
+    for w in word_list:
+        if w in word_freqs:
+            word_freqs[w] += 1
+        else:
+            word_freqs[w] = 1
+    return word_freqs
+
+def sort(word_freq):
+    return sorted(word_freq.iteritems(), key=operator.itemgetter(1), reverse=True)
+
+def top25_freqs(word_freqs):
+    top25 = ""
+    for tf in word_freqs[0:25]:
+        top25 += str(tf[0]) + ' - ' + str(tf[1]) + '\n'
+    return top25
+
+#
+# The main function
+#
+pa = TFPassiveAggressive(None)
+pa.bind(get_input).bind(extract_words).bind(remove_stop_words).bind(frequencies).bind(sort).bind(top25_freqs).printme()
+