Alexander Index: design and concepts



This page explores in more detail key features and design concepts of the Alexander Index ("the Index", for short).

Index or concordance?

Briefly, the differences between an index and a concordance are that

  • a concordance aims to record all the occurrences of a word - or possibly a phrase - within a text; it's relatively simple;
  • a good index has been decribed as “a structured sequenceresulting from a thorough and complete analysis of textof synthesized access points to all the information contained in the text ... a network of interrelationships ultimately an interface between the author and the reader"1.

A concordance can be produced more easily through software, whereas an index is more demanding because

  • it requires a human to understand the text in sufficient depth to know what the appropriate "access points" are;
  • it often requires that the sense of a statement is boiled down to its essentials in order to produce a usable index entry;
  • attention may need to be given to the order of words in the phrase so that the relevant one appears first.

Both index and concordance have their value. In an index aimed at a more informed readership particular turns of phrase can be interesting, not just the underlying thought: it will be the aim of the Index to capture those individual turns of phrase, which may disclose important subtleties in Alexander's thinking.  In cases such as these, the Index will operate as a concordance. 

On the other hand, given the size of the Index, it is important that entries are reasonably specific in order to direct the enquirer to the most relevant text.  Simply registering all occurrences of the word "inhibition" for example, whilst valuable for some forms of textual analysis, is not likely to provide pointers that are suitable for those wishing to locate or discover particular contexts for the usage of the word2.  The index has to point towards identifiable concepts, ideas and topic areas not just list words and phrases that may be used in very different ways.  In conclusion then: the primary function of an index has to be maintained. 


We are very fortunate that Jean Fischer has provided us with his indexes, in digital form, of those books authored by Alexander that are published by Mouritz (Articles and Lectures excepted for now).   The digital files are being processed to convert them into a form suitable for uploading to the ASO Alexander Index: this is a demanding piece of work but not nearly as demanding as it would have been to start from scratch. 

The index of The Use of the Self has been created manually using as a basis the index of the printed book, as prepared by George Trevelyan and Gurney MacInnes (both students on Alexander's first training course).  But the entries have been systematically reviewed and now include numerous additions and amendments.

These two sources are somewhat different in their approach to indexing. Jean Fischer's entries tend to be briefer compared with those in the Use of the Self  but they record individual occurrences of words and phrases more systematically: they are towards the concordance end of the spectrum. The source index for the Use of the Self, on the other hand, explores concepts perhaps more deeply but lacks the systematic recording of word-occurrences and highlighting of phrases. Both require some reworking to reflect the preferred approach of the ASO project.

The net result of using different sources is that the Alexander Index is currently inconsistent in its approach. This is regrettable but not a deal-breaker.


The index consists of a single database table containing series of index records of identical format.

Each index record contains a main entry, a subentry (optional) and a subsubentry (also optional): all three of these together comprise the Index Entry. When listed in tabular form the separate components of the entry are run together, separated by colons.

Each index record contains a locator field and either 

  1. the locator field will contain one or more page references where the index entry can be found; or
  2. the locator field will be blank and instead a subentry or subsubentry will contain a cross-reference (containing see... or see also...) pointing to another index entry.

Each index record is linked to a single Text record in the Bibliography database, details of which can be found here.

It follows from the above that a single word or phrase that has been indexed can have multiple index records, one for each text in which the word and phrase can be found. 




The challenge of multiple texts

The size of the Index stems from the fact it references multiple texts.  This imposes a need for precision in index entries, so they are suitably specific

Checkng of index entries prior to uploading  to make sure they are sufficientely specific.

As new entries are added to the Index, regular reviews will be required to see if some entries are becoming unmanageably congested (too many page numbers and texts) and would warrant re-indexing with more refined terms.

Multiple texts means multiple index authors who may have selected different terms to index: there is a need for consistency.

Adjudication is necessary to select the most appropriate terms to index and amend other entries to make them consistent. For example "acts", "actions" and "activity": should these be indexed separately or treated as essentially the same?

A judgement has to be made as to how significant the original usage was and whether to retain it as a separate entry; otherwise it will appear as a cross-reference.

Cross-references ("see ....", "see also...") need to refer across all texts A "virtual" text has been created to hold all cross-references across the entire Index, so that cross-references need only occur once, rather than once for each text.


The challenge of history

Terminology changed over time; but sometimes the concepts were the same. Addressed via appropriate use of cross-referencing, though normally text is indexed according to the vocabulary used with see also entries pointing to other usages.


The challenge of multiple authors

With the large volume of index entries covering a wide range of texts, it will be important to establish who is the author of any particular section of text.  In the indexes of the print versions such differentiation is not available. Books have been split down into texts according to authorship. So, for example, in the bibliography referenced by the Index The Universal Constant in Living has separately identified texts for George Coghill's Introduction and Jean Fischer's footnotes (as well as Alexander's main text and several others whose texts appear in the book).

The challenge of multiple editions

There are multiple editions of Alexander's works, many with different page numbering, some with different texts.

Indexing multiple editions of the works would mean a proliferation of index entries, many identical other than the page number.

As an experiment the main text of the 1946 Edition of The Use of the Self has been indexed alongside the current (1985 onwards) edition.  But in general, the works indexed are the current, scholarly editions ie those most easily available now and with the richest content.


1. Nancy C. Mulvany, Indexing Books, 2nd Edition, University of Chicago Press, 2005.

2. Moreover, those involved in detailed textual analysis will probably find ways to digitise the books in order to pursue their activities.