This and That and The Syntax of Concealment

How can you hide secret information in plain text? You could try non-obvious manipulations of this and that. First, some background …

Dr-MurphyDR-vogelSteganographers have been endeavouring to devise methods for hiding information since at least 1499 (and probably before that). A subset, known as Stegotext, hides info in apparently normal text – for instance by manipulating the formatting in subtle ways. But there are disadvantages with this method, if the text is copied and re-formatted the message is lost.

A new (2007) method is proposed which relies instead on covert manipulation of complementisers (e.g. ‘that’) and/or relativisers (e.g. ‘that’). It was devised by Dr. Brian Murphy, who’s now at Queen’s University, Belfast, and Dr. Carl Vogel of the Computational Linguistics Lab, O’Reilly Institute, Trinity College, Ireland. The new method relies on the addition, deletion, and/or swapping of words such as ‘this, that, or ‘who’ without in any way altering the sense of the text. The researchers say their method has a success rate of 96% and achieves a “bandwidth of 0.3 bits per sentence.” (or, as Improbable prefers to put it, 1 bit every three sentences, or so.)


The syntax of concealment: reliable methods for plain text information hiding’ (in Proceedings of SPIE – The International Society for Optical Engineering. Vol. 6505, 2007.)


The syntax of concealment: reliable methods for plain text information hiding’  (in Proceedings of Security, Steganography, and Watermarking of Multimedia Contents IX, 2007.)

BONUS QUIZ: Could we have we covertly hidden any information in this post?

Improbable Research