After quoting several pieces by a German author, whose names contains an Umlaut, I get this when generating the bibligraphy:

Habermas, Jürgen. Erläuterungen zur Diskursethik. 5. Aufl. Frankfurt am Main: Suhrkamp, 1991.
———. Moralbewußtsein und kommunikatives Handeln. Frankfurt am Main: Suhrkamp, 1983.
Habermas, Jürgen. Theorie des kommunikativen Handelns. Bd. 1. 2 Bd. Frankfurt am Main: Suhrkamp, 1981.
———. Theorie des kommunikativen Handelns. Bd. 2. 2 Bd. Frankfurt am Main: Suhrkamp, 1981.
Habermas, Jürgen. „Three Normative Models of Democracy“. Constellations 1, Nr. 1 (1994): 1–10.

Of couse, what I wanted was this:

Habermas, Jürgen. Erläuterungen zur Diskursethik. 5. Aufl. Frankfurt am Main: Suhrkamp, 1991.
———. Moralbewußtsein und kommunikatives Handeln. Frankfurt am Main: Suhrkamp, 1983.
———. Theorie des kommunikativen Handelns. Bd. 1. 2 Bd. Frankfurt am Main: Suhrkamp, 1981.
———. Theorie des kommunikativen Handelns. Bd. 2. 2 Bd. Frankfurt am Main: Suhrkamp, 1981.
———. „Three Normative Models of Democracy“. Constellations 1, Nr. 1 (1994): 1–10.

The whole thing appears to be related to the entries using different encodings; the following Python output corrobaretes this suspicion:

Python 2.7.1 (r271:86832, Jul 31 2011, 19:30:53) 
[GCC 4.2.1 (Based on Apple Inc. build 5658) (LLVM build 2335.15.00)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> len("Habermas, Jürgen") # First occurrence
17
>>> len("Habermas, Jürgen") # Second occurrence
18
>>> len("Habermas, Jürgen") # Third occurrence
17

The difference between the first and third occurence is somewhat curious:

>>> [f for f in "Habermas, Jürgen"] # First occurrence
['H', 'a', 'b', 'e', 'r', 'm', 'a', 's', ',', ' ', 'J', '\xc3', '\xbc', 'r', 'g', 'e', 'n']
>>> [f for f in "Habermas, Jürgen"] # Second occcurrence
['H', 'a', 'b', 'e', 'r', 'm', 'a', 's', ',', ' ', 'J', 'u', '\xcc', '\x88', 'r', 'g', 'e', 'n']
>>> [f for f in "Habermas, Jürgen"] # Third occurrence
['H', 'H', 'a', 'b', 'e', 'r', 'm', 'a', 's', ',', ' ', 'J', '\xc3', '\xbc', 'r', 'g', 'e', 'n']

I've got no idea where the second "H" in the third occurrence is coming from. Thanks for looking into this.