Consider the following saturated chain in the Young's Lattice: $$ \phi \subset \lambda^{1}\subset \lambda^{2}\subset \cdots \subset\lambda^{n}$$ where $\phi$ is the empty partition and $\lambda^{i}$ is a partition of $i$, denoted $\lambda^{i} \vdash i$, for $i=1,2,\ldots ,n$.

I am going to consider the Ferrers diagram of partitions in French notation, and reference the cells of the Ferrers diagram by the pair (row,column).

With the aforementioned saturated chain, associate a column growth word $w=w_1w_2\ldots w_n$ in the following manner: $$w_i= \text{ column of the cell } \lambda^{i}/\lambda^{i-1}$$ Starting from the empty tableau, do jeu-de-taquin slides starting from the outer corner as given by the word $w$ to recover a standard Young tableau(SYT) of shape $\lambda^{n}$. This gives one a(nother) bijection between saturated chains in Young's Lattice and SYTs.

$\textbf{Question}$: Is it possible to predict what the SYT associated with a saturated chain will be, just by the knowing the column growth word, so as to avoid making $n$ jeu-de-taquin slides?

What follows is what I have been doing. I would like a reference where this has already been dealt with, or is implicit (since I am pretty sure this has to be in the literature somewhere).

1) Reverse the word $w$ to obtain $w^r$.

2) Standardise the word $w^r$ to obtain a permutation $\pi$.

3) $\pi \longleftrightarrow (P,Q)$ via the RSK correspondence.

4) $Q^{t}$ (the transpose of the tableau, that is) is the SYT we are looking for.

Example for making the notation clear: Consider the maximal chain $\phi \subset (1)\subset (1,1)\subset (2,1)\subset (2,1,1)\subset (2,2,1)$. Then the column growth word associated with this chain is $w=11212$. Then $w^r=21211$ and the standardisation gives the permutation $41523$ in single line notation.


I believe that this is proved by E. Gansner, On the equality of two plane partition correspondences, Discrete Math. 30 (1980), 121-132.

  • $\begingroup$ Thanks for the reference, Prof. Stanley. Let me look at the paper in detail before I accept the answer. From a first glance, surely the paper seems to be doing the sort of thing I want. $\endgroup$ Jun 25 '12 at 6:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.