As detailed in the previous post, I was awarded the Wilhelm von Humboldt prize 2014 for my dissertation at this year's annual meeting of the DGfS (German Linguistic Society) in Marburg. During the ceremony, I had to present the dissertation in 10 minutes to an audience of linguists of all flavors. First I planed to really present the work, but quickly realized that this was not only a hopeless endeavor, but also the wrong take on such a presentation. I therefore decided to focus on the people and ideas that influenced and shaped me as a linguists and directly or indirectly had an influenced the final shape of my dissertation. Since the slides of the presentation were in German, here comes an English version in form of a blog post.


Use-conditional meaning: Biography of a dissertation


The dissertation in buzzwords

Formal pragmatics — Applying the formal tools of semantics to traditional pragmatic topics.

Use-conditional meaning — Some expressions of natural language impose conditions on their felicitous use, instead of contributing to the truth-conditioans of a sentence.

Hybrid semantics — In order to capture the entire meaning of an expression, one needs to take into account the truth- and use conditional dimensions of meaning.

Multidimensionality —  The meaning of natural language expressions consist of (at least) this two dimensions , which are independent of each other but interact compositional with each other.

Compositionaliy — In contrast to some previous multidimensional approaches, the formal system developed in the thesis is strictly compositional and uses just two compositions rules.

Modal particles — Besides classic expressives like expressive adjectives (damn etc.), one of main empirical domains studied in the dissertation are modal particles in German.

Sentence mood — his also leads to a fresh perspective on sentence mood, under which mood operators are understood as use-conditional expressions. This makes it possible to represent them explicitly without inheriting the problems of the »performative hypothesis«.

Extendability – The system is general enough that it can be extended by additional dimensions (e.g. for appositives or focus) and can be applied to other field of linguistics investigations (language change, typology).


But today: people before content


  • My main teacher during my MA program at Mainz
  • Got me into modal particles ...
  • ... and sentence mood
  • Taught me about pragmatics, and especially, semantics vs. pragmatics.
  • Introduced me to issue in the philosophy of language (what lead me to study philosophy as a minor in my MA)

To understand a proposition means to know what is the case if it is true.
— Wittgenstein 1922: §4.023

➢ Semantic Theory of Meaning

… the meaning of a word is its use in the language.
— Wittgenstein 1953: § 43

➢ Pragmatic Theory of Meaning

Is one right?     (Spoiler alert: No, both!)     


  • Together with Jörg sparked my interest in linguistics with his Intro to Linguistics class.
  • Later, he recommended me Chris Potts's just appeared book.
  • Also pointed me towards Horst Lohnstein's semantics text book.
LCI.jpg
  • Conventional implicatures made me curious, as I heard they were a myth.
  • Expressives also seemed to be relevant for my seminar paper on ethical datives.
  • But ?λx??⟨e, t⟩?∀y?●????
BuchHorst.jpg
  • I got my semantic »basic training« from this book.
  • It was hard to get at that time, only 2nd hand. (There is a new edition now.)
  • I worked through this book with goal of understanding Potts's book. 

Chris Potts (now!)

  • Expressions of natural language can carry more than one meaning.
  • In case of conventional implicatures, those dimensions of meaning are independent of each other.

Example: appositives

  • Daniel, a linguist, writes a blog post.
  • ⟨ Daniel writes a blog post, Daniel is a linguist ⟩

 

  • Potts develops also a formal System to model the different dimensions and their interaction.
  • If one can have more than dimension of meaning, maybe the Wittgenstein can be unified with Wittgenstein after all?
  • But how can a unified formal approach be possible, given that pragmatic use-theories of meaning are mostly construed in an anti-formal manner?

Anonym.jpg
  • Somebody provides me with a copy of Kaplan's The meaning of ouch and oops (1999), the infamous underground paper, to which Potts 2005 drew my attention.
  • This helpful person therefore must remain anonymous.

For certain expressions of natural language, a correct Semantic Theory would state rules of use rather than something like a concept expressed.
— David Kaplan (1999: 6)

Truth-conditions 

  1. »Snow is white«
  2. ist true,
  3. iff snow is white.

Use-conditions

  1. »Oops«
  2. is felicitously used,
  3. iff the speaker observed a minor mishap.
  • In both conditions, an expression is connected with a condition that captures its meaning.

  • What differs is the kind of connection (»mode of expression«).

  • These conditions can be the case or not.

  • This makes them propositional → Standard semantic toolbox available! 


Potts + Kaplan = Hybrid Semantics

  • Assuming a multidimensional semantics à la Potts, we can view the meaning of a natural language expressions as the combination of their truth- and use-conditional content.
  • We can have our cake and eat it, too!

Truth- and use-conditions (here in the case of modal particles)

  1. »Es regnet ja« (it's raining MP) is true, iff it is raining.
  2. »Es regnet ja« (it's raining MP) is felicitously used, iff it is common knowledge that it is raining.
  • Expressions with both meaning dimensions are hybrid  expressions.

  • Hybrid semantics: 1, ✓⟩   ⟨1, ✗ ⟩   ⟨0, ⟩   ⟨0✗⟩

  • Many complex expressions are hybrid (that damn Daniel), but there also lexically hybrid expressions (Kraut).


  • From Ede, I got my semantic »master class«.
  • Lead me to think more about compositionality.
  • Spark my interest in the important role of contexts.
  • Became the principle supervisor of the thesis.
  • Answered thousands of questions and inspired me on even more occasions.
  • What is the nature of the use-values ✓and ✗?
  • Since, as seen above, use-conditions can be the case of not, the notions of truth sneaks back into the theory.
  • If an ordinary truth-conditional proposition is understood as a set of worlds, a use-conditional proposition can be conceived as a set of contexts (in order to be used felicitously, there must be an actually utterance and hence context).
  1. ⟦Snow is white⟧-t = {w: snow is white in w}
  2. ⟦Oups!⟧-u = {c: S(c) observed a minor mishap on w(c)}

 

  • Potts's system, however, involves some compositionally problems, which are based on the fact that expressions are only sporadically multidimensional.
  • In contrast, the framework developed in the dissertation is based on consequent multidimensionality: every expression exhibits both a truth- and a use-conditional dimension (and an intermediate dimension for compositional bookkeeping).
  • These dimensions are calculated side by side during semantics derivations.
  • So-called lexical extension rules ensure that »one-dimensional« expressions like house or pus get equipped with the necessary dimension to avoid to blow up the lexical entries of such expressions.
 How an actual derivation actually looks like in my thesis. I am really happy that the jury – in spite of this – especially mentioned the accessibility of the my dissertation.

How an actual derivation actually looks like in my thesis. I am really happy that the jury – in spite of this – especially mentioned the accessibility of the my dissertation.


To conclude: modal particles and sentence mood again

  • A use-conditional and multidimensional analysis of modal particles cannot only account for their semantic status, but is also able to derive many of their syntactic particularities.
  • The notion of sentence mood also receives a fresh analysis, as sentence mood operators are also treated as use-conditional expressions.
  • This allows one to represent them explicitly in the semantics without leading to wrong truth-conditions.
  • The interaction between modal particles and sentence mood can then also be modeled in the use-conditional dimension.
  • A sentence involving various use-conditional expression can therefore lead to complex use-conditions.
  1. ASSERT(Der verdammte Daniel trägt halt vor)    “That damn Daniel is MP talking.”
  2. ⟨ Daniel is talking, the speaker wants that the hearer knows that Daniel is talking & the speaker has a negative attitude towards Daniel & one cannot do anything about Daniel being talking ⟩

  • Through his papers, I learned that there is much more to expressive meaning than was accounted for by Potts 2005.
  • That thought me to not fear take an existing (and established) framework apart in order to build something new out of it.
  • He became the second supervisor of the thesis.

From him, I learned that one should respect old work and the early pioneers an that there are still  a lot of hidden ideas waiting to be uncovered.

Lead me to think more about what speech acts are and how they relate to sentence mood.

He became the third supervisor of the thesis.


  • Beside those that are explicitly mentioned on this page, there are many more persons that help and inspired me on my way from my first contact with linguistics to putting together the final version of my dissertation.
  • A revised version of the dissertation will published by OUP (probably in 2015).
  • Who want to get the original version of the thesis, can contact me.

Multidimensional Thanks!

Comment