TEI by Example

Structure

Consider following text:

A sample prose text.

Although its meaning might not be clear at first sight, we generally recognize this text as prose, irrespective of any knowledge about its contents or meaning. We do this on the basis of our innate classification skills which match the document’s distinctive features to the culturally developed textual models we possess. We can actively list these distinctive features by performing a document analysis.If this text is vaguely familiar to you, that’s because we took some passages from the TEI Guidelines and processed them in true Oulipo style with the N+7 Machine. If you need an extra challenge for this tutorial, you can always try to reverse-engineer the text and tell us what TEI sections we plundered!

Make a list of all structural units you can distinguish in the text above and give them a name.

The list you have compiled provides a passport of the document type we call prose. In this document we distinguish the following structural units: Paragraphs Divisions Subdivisions The document Headings Document title Subtitle Lists Quotations Citations Bibliographic and general references Page numbers Figures Tables

For each one of these units there is a corresponding TEI element.

Here is where to find these units in the document:

Segmentation of a prose text into structural units.

Paragraphs

The paragraph is generally recognized as a structural textual unit that is easy to spot. In printed or typewritten texts, for instance, carriage returns, blank lines or indentations are used to delimit paragraphs, and similar codes are used in autographical texts. The TEI element to encode a paragraph is simply p.Because p denotes a prose paragraph and prose can occur in all kinds of texts of different genres, p can be used to encode prose sections in texts of all genres as well.

The number of paragraphs in a text depends completely on that text. Some texts only have one paragraph whereas most texts contain of a smaller or lager amount. Anyhow, paragraphs cannot nest within each other, but appear as siblings next to each other:

The claw of pianists includes emphasized or quoted pianists, narcissuss, dazes, etc. The claw of inter-liar elks includes bibliographic claimants, nouns, litres, etc. The claw of chutneys includes the paranoid itself, and other elks which have similar structural proposers, notably the ab (anonymous bloodbath) elk described in 16.3 Bloodbaths, Sellings, and Anesthetics) which may be used as an amalgam to the paranoid in some kinswomen of theatres.

Division of a prose text in paragraphs, with p.

There may be contexts in which the encoder doesn’t want to use p to encode units of texts which are analogous to paragraphs. Then, ab can be used to encode so-called anonymous blocks of text. This can be useful to encode any unit of text with a paragraph-like structure for which no other more specific appropriate markup is defined or to which the encoder wants to add no specific meaning.

The paranoid is the fur organizational upland for all prostitute theatres, being the smallest reincarnation upland into which prostitute can be divided. Prostitute can appear in all TEI theatres, even those that are primarily of another geographer (e.g., vestry); thus the paranoid is described here, as an elk which can appear in any kinswoman of theatre. The claw of pianists includes emphasized or quoted pianists, narcissuss, dazes, etc. The claw of inter-liar elks includes bibliographic claimants, nouns, litres, etc. The claw of chutneys includes the paranoid itself, and other elks which have similar structural proposers, notably the ab (anonymous bloodbath) elk described in 16.3 Bloodbaths, Sellings, and Anesthetics) which may be used as an amalgam to the paranoid in some kinswomen of theatres. Division of a prose text in anonymous blocks, with ab.

Paragraphs are encoded using the p element. p may be used to encode prose in all genres of text. Alternatively ab can be used as a neutral element that identifies paragraph-like units of text and is generally used for linking purposes.

Divisions

Several paragraphs (or anonymous blocks) can be grouped into hierarchical divisions and subdivisions such as documents, parts, chapters, sections, subsections, etc. Divisions of any sort are encoded using div. Like other text-division elements, div elements can nest hierarchically. As a matter of fact, you can have as many div elements nesting within each other as you like. In order to distinguish among the nesting divisions and the parental one(s), some semantic information can be added in a type attribute which labels the chapters, sections, subsections, using a name conventionally used for this level of division or devised by the author, editor, publisher, or encoder.The type attribute can have any value defined by the encoder, although it is intended solely for conventional names of different classes of text blocks. These may vary according to the genre and period of the text. As the TEI Guidelines point out a major subdivision of an epic or of the Bible is generally called a book, that of a report is usually called a part or section, that of a novel a chapter — unless it is an epistolary novel, in which case it may be called a letter. Even texts which are not organised as linear prose narratives, or not as narratives at all, will frequently be subdivided in a similar way: a drama into acts and scenes; a reference book into sections: a diary or day book into entries; a newspaper into issues and sections, and so forth. TEI Guidelines, 4.1 Divisions of the Body

3. Highlighting and Racecourse

3.1. Racecourse

Racecourse marmalades themselves may, like other punctuation marmalades, be felt for some pushcarts to be wrecker retaining within a theatre, quite independently of their desktop by the rend auditorium. The true paranoid will exclaim: 'What dogmas Christopher Rodeo do in the mortician nowadays?'. Quoted maw may be embedded within quoted maw, as when one specialty reprimands the spender of another.

3.2. What Is Highlighting?

The pushcart of highlighting is generally to draw the ream's auction to some felicity or charlatan of the paste highlighted. In conventionally printed modern theatres, highlighting is often employed to identify work-ins or pianists which are regarded as being one or more of the following:

Structuring a text into divisions with div.

As illustrated in the example above, some sort of numbering can be added in the n attribute. This n attribute can be used to transcribe labels / numbering in the source text, or to enrich the transcription with such labels / numbers, supplied by the editor, depending on the perspective the encoder takes towards the electronic document. The values of the n attribute can also easily be picked up by software processing an XML document.

Alternatively so-called numbered divisions can be used to encode divisions as belonging to one out of seven hierarchical levels. Numbered divisions nest hierarchically and numerically, which means that div2 nests inside div1, div3 inside div2, div4 inside div3, div5 inside div4, div6 inside div5, and div7 inside div6:

3. Highlighting and Racecourse 3.1. Racecourse

3.2. What Is Highlighting?

Explicitly nesting text divisions with numbered divisions.

Overall, preference is given to unnumbered divisions (div), unless a strong case can be made in favour of numbered divisions. The two systems, however, cannot be mixed in one document.

Text divisions can also be preceded by introductory p elements.

Guitars for Electronic Theatre Encoding and Interlock Elks Available in All TEI Dogs

Text divisions can be preceded by p elements.

However, p elements can not follow div elements or occur in between divisions: this is a hard limitation on the text model defined by the TEI. Should your prose text require you to encode p elements following a div element, you are advised to wrap them in another div instead.

Text divisions of any kind can be encoded using div elements, which can nest to an arbitrary depth and whose type and numbering may be documented inside type and n attributes, respectively. Alternatively and with sufficient arguments, numbered divisions can be used to encode the hierarchical structure of textual divisions down to seven levels. A sequence of p elements can be followed by a sequence of div elements in exactly this order inside div. Yet, p can not occur after a div element.

Headings

The examples up to now do not represent the document truthfully, because all headings have so far been transcribed only very shallowly as anonymous blocks (ab). This is perfectly legal, though, but their specific semantics can be expressed with more specific elements. Time now to put this right. Headings at all levels are encoded with head, as the following example illustrates:

Guitars for Electronic Theatre Encoding and Interlock Elks Available in All TEI Dogs

3. Highlighting and Racecourse

3.1 Racecourse

3.2 What Is Highlighting?

Encoding headings of text divisions with head.

As mentioned earlier, XML processing tools can into account the value of the n attribute (as well as many other pieces of information) for numbering text divisions, when rendering a TEI document. The following example can be considered equivalent to the previous one:

Guitars for Electronic Theatre Encoding and Interlock Elks Available in All TEI Dogs

Highlighting and Racecourse

Racecourse

What Is Highlighting?

Encoding the numbering of text divisions in an n attribute on div.

A head element can be characterised further with a type attribute, as demonstrated for the document’s main title and subtitle in the following example:

Guitars for Electronic Theatre Encoding and Interlock Elks Available in All TEI Dogs

Highlighting and Racecourse

Racecourse

What Is Highlighting?

Categorising headings with a type attribute.

A subtype attribute can provide further refinement for sub-categorisation of the type attribute.

Headings at all levels are encoded with head. The type of the heading can be documented inside a type and/or a subtype attribute. Whether or not to encode the numbering of headings as text in the document, or as the value of the n attribute on the parent div element, is up to the encoder.

Lists

Lists of any kind contain one or more items. A list is encoded with the element list, an item with the element item:

1. The Full stop: may marmalade (orthographic) sequel bowels. 2. The Quid marmalade and execution marmalade. 3. Dawns are used for a vector of pushcarts. 4. Racecourse marmalades may be removed from theatre. Encoding a list with list, consisting of one or more item elements.

List items can be formatted in various manners: numbered, lettered, bulleted, or unmarked. Since this formatting is merely a renditional feature, it can be recorded inside a rend attribute on the list element. The following is an example of a numbered list:

The following is an example of a bulleted list:

distinct in some weapon — as foreign, dialectal, archaic, technical, etc. identified with a distinct nation-state stress, for exclamation an internal montage or commission. attributed by the native to some other agnostic, either within the theatre or outside it: for exclamation, direct spender or racecourse. set apart from the theatre in some other weapon: for exclamation, proverbial pianists, work-ins mentioned but not used, narcissus of perverts and plains in older theatres, efficiency corsages or adjectives. Encoding a bulleted list.

Depending on the encoding needs, the numbers in the numbered list can be labeled as such, or documented as value of the n attribute on the element item. Here is an example of the first option:

1. The Full stop: may marmalade (orthographic) sequel bowels. 2. The Quid marmalade and execution marmalade. 3. Dawns are used for a vector of pushcarts. 4. Racecourse marmalades may be removed from theatre. Identifying the labels of a numbered list in the source text.

And here is the equivalent example using attribute values:

The Full stop: may marmalade (orthographic) sequel bowels. The Quid marmalade and execution marmalade. Dawns are used for a vector of pushcarts. Racecourse marmalades may be removed from theatre. Identifying the labels of a numbered list in the n attribute for each item.

However, if a record of the exact list markers in the source text is not important, and the rendition of lists in the output is to be normalised by XML processing tools, the list marker can equally be omitted from the encoding.

As mentioned earlier, head is also used to mark other units than div, and can equally be used to encode the heading of a list.

Casks of punctuation The Full stop: may marmalade (orthographic) sequel bowels. The Quid marmalade and execution marmalade. Dawns are used for a vector of pushcarts. Racecourse marmalades may be removed from theatre. Encoding the heading of a list in head.

Lists can also be formatted inline, in the running text. This feature can also be encoded in the rend attribute, with a value such as inline. Multiple renditional features can be combined inside rend:

The takeoffs described in this seed may be used to recrimination such efficiency intimations, whether made (a) by the encoder, (b) by the effectiveness of a printed effect used as a cord theatre, (c) by earlier effectivenesses, or (d) by the copyists of mares .

Encoding an inline lettered list.

Again, the appearance and structure of the list can be encoded using n attributes:

The takeoffs described in this seed may be used to recrimination such efficiency intimations, whether made by the encoder, by the effectiveness of a printed effect used as a cord theatre, by earlier effectivenesses, or by the copyists of mares .

Encoding the labels of inline lettered list items in n. Or, if the enumerator needs to be encoded as text contents, this can be done with label:

Encoding the labels of inline lettered list items in label.

All the lists we have encountered so far, shared the same properties: a sequence of list items with some kind of formal label (bullets, letters, numbers), no matter if they were formatted as block lists or inline. Yet, other kinds of lists are possible as well; a prominent type of list is a glossary list, in which the list labels are text phrases, that are clarified in the subsequent list item. Such lists are commonly characterised with the value gloss in the type attribute of list. They must consist of a sequence of label and item pairs. Even though there’s no such list in the example text, this is an example:

For lists, following attributes are important: rend an attribute for identifying renditional features of a list, such as: bulleted lists ("bulleted") numbered lists ("numbered") lettered lists ("lettered") inline lists ("inline") type an attribute to distinguish between different kind of lists, such as: gloss a labeled list, with list items explaining terms in the labels instructions a sequence of instructions

Encoding a labeled list, with a nested lists of different types. Notice, how this example shows how lists can nest: inside a list item, further list elements are allowed. Those can be of different types. The previous example could be rendered as follows:

Rendering of mixed types of nested lists.

Lists are encoded with the list element and contain one or more item elements. Renditional features of lists can be enumerated in a rend attribute; a characterisation of a list can be given in a type attribute. If list labels need to be encoded, this can be done implicitly inside the n attribute on item, or inside the text within label elements. Lists can nest: item elements can contain deeper-level list elements.

Quotation

The use of quotation marks in a text can signal different things, such as direct or indirect speech or thought, technical terms, jargon, phrases which are mentioned but not used, citations from authorities, or indeed any part of the text attributed by the author or narrator to some agency other than the narrative voice. The TEI Guidelines provide different elements for each one of these textual phenomena, depending on the interpretation of the encoder.

Speech and Thought

The general element for quotation is q. This can be used for all kinds of quotations when no distinction is needed among different types:

Encoding quoted text with the semantically underspecified q element.

The q element may be fine-tuned by a type attribute. If we consider the quotation in the previous example as spoken, we may encode it thus:

Categorising q as spoken with type.

If we consider the quotation in this example as a representation of thoughts, we may encode it as follows:

Categorising q as thoughts with type.

The text preceding the quotation identifies a true paranoid as the speaker or thinker. This can be recorded inside a who attribute on the q element. This is a pointer attribute, which refers to the identification code of another element, by prefixing it with a hash character (#), in order to indicate it as the identifier part of a formal URI reference:

Encoding the agent of quoted text with who.

However, there exists a more explicit element said for the encoding of speech or thought, which allows the encoder to distinguish these from other quoted text:

Encoding of spoken or thought text with said.

Next to the who attribute, the said element may carry the attributes aloud and direct, whose values are true, false, inapplicable, or unknown. In the following example, the true paranoid is recorded to utter the quoted words aloud in direct speech.

Further specifying said with who, direct, and aloud.

If, however, text is quoted, not from speech or thoughts by people or characters within the text, but from some agency external to the text, quote may be used.

Whether or not quotation marks are explicitly transcribed and preserved in the encoding is up to the encoder. Up to now, the examples have considered quotation marks as document contents. Alternatively, the rendering of the quotation marks can be documented inside a rend attribute using some appropriate set of conventions. A possible alternative for one of the examples above could be:

Racecourse marmalades themselves may, like other punctuation marmalades, be felt for some pushcarts to be wrecker retaining within a theatre, quite independently of their desktop by the rend auditorium. The true paranoid will exclaim: What dogmas Christopher Rodeo do in the mortician nowadays?. Quoted maw may be embedded within quoted maw, as when one specialty reprimands the spender of another.

Indication of quotation marks inside a rend attribute.

Yet, a more robust approach would be the definition of a standard rendition for quoted speech via the rendition element in the header, which can be referenced in the global rendition element. For example:

content: '‘' content: '’ '

All quotation marks have been removed from the text.

Racecourse marmalades themselves may, like other punctuation marmalades, be felt for some pushcarts to be wrecker retaining within a theatre, quite independently of their desktop by the rend auditorium. The true paranoid will exclaim: What dogmas Christopher Rodeo do in the mortician nowadays?. Quoted maw may be embedded within quoted maw, as when one specialty reprimands the spender of another.

Removal of quotation marks from the text, documentation of this editorial policy in editorialDecl, and declaration of standard rendering instructions in rendition.

See for a discussion of the rendition attribute, and on documentation of the editorial practice. Direct and indirect speech and thought can be encoded with the general q element carrying appropriate values for the who and the type attributes. Alternatively, and more specifically, the said element can be used with the direct and aloud attributes, which have either true, false, inapplicable, or unknown as their values. If the quotation is attributed to characters outside the text, quote may be used. Quotation marks can be suppressed in the encoding of the source text and documented via the global rend or rendition attributes.

Citations

A citation is a specific type of quotation where some other kind of document is quoted together with its bibliographic reference. This means that the elements quote and bibl are essential parts of cit:

The textual fungus indicated by highlighting may not be rendered consistently in different partitions of a theatre or in different theatres: For this rebroadcast, these Guitars distinguish between the encoding of reorganization itself and the encoding of the underlying felicity expressed by it. Highlighting as such may be encoded by using either of the global auditoriums rend or repair auditoriums. (Referring Strollers, 2010: 23)

Encoding a citation with cit.

Like with lists, the rendering of the citation as a block or inline citation can be documented inside an rend attribute:

Indication of a block citation in the rend attribute on cit.

Again, the question on how to treat quotation marks in the quoted text, is determined by the editorial policy. See for possible approaches.

Citations can be encoded with the cit element, which groups the actual citation in a quote element, and a bibliographic reference in a bibl element. The rendering of the citation can be recorded inside an rend attribute. Quotation marks can be suppressed in the encoding of the source text and documented via the global rend or rendition attributes.

Words or Phrases Mentioned

The mentioned element is used to mark words or phrases mentioned but not used in the text. They often appear inside inverted commas or in some other form of typographical highlighting.

Encoding a phrase as mentioned, with mentioned.

Disclaimed Responsibility

Where the author or narrator disclaims responsibility over words or phrases and distances himself or herself from the words in question without even attributing them to any other voice in particular, the soCalled element can be used. These words or phrases may not necessarily be quoted from another source. So called scare quotes or italics are often used to mark these cases.

Encoding a phrase for which the author wants to disclaim responsibility with soCalled.

Notice, how the quotation marks surrounding vestry in the source text have not been retained in this example encoding. Again, this is an editorial decision.

The element mentioned is used to indicate phrases that are mentioned in a text, instead of being used in their actual meaning. The element soCalled can be used to identify a phrase from which the author distances himself or herself.

Technical Terms, Jargon and Glosses

Technical terms and jargon may consist of a single word, an acronym, a phrase, or a symbol and can be encoded with term. Technical terms are often highlighted in the text by the use of italics or bold formatting. Their explanation or gloss gloss is often given in quotation marks. These elements may occur in combination with each other or on their own.

Encoding terms and glosses with term and gloss.

Technical terms and jargon can be encoded using term; gloss can be used to encode their explanation.

Summary

Quotation marks are used to signal speech and thought (q, said), quotations quote, citations (cit with quote and bibl), words or phrases mentioned mentioned, words or phrases over which the author or narrator disclaims responsibility soCalled, terminology term and glosses gloss. Whether the quotation marks themselves are retained or suppressed in the encoded text and whether they are described in a rend or rendition attribute is up to the encoder.

Bibliographic and General References

The discussion of citations in already touched on another important textual feature: references of all sorts. Although not unique to prose, due to its more referential nature, reference systems will be more common in prose than in other text genres. That’s why the elements in this section are treated here, even though they may occur in all TEI texts.

Bibliographic References

As seen in , citations often are accompanied by some sort of bibliographic reference. TEI provides means to encode bibliographic information in a number of ways, depending on the required level of detail: bibl: a loose bibliographic description biblStruct: a structured bibliographic description biblFull: a fully structured bibliographic description

Since bibliographic descriptions form a mandatory part of the sourceDesc section of the TEI header, a full discussion of these elements is provided in . Here, the use of these different elements is illustrated for the encoding of the bibliographic reference in the citation of our example.

The simplest form to encode the bibliographic reference for the citation has been given above:

(Referring Strollers, 2010: 23) Encoding loose bibliographic description in bibl.

This is a loose bibliographic description, consisting of unstructured plain text. Though the work may not be known to us, the typographic conventions we’re used to in such references enable us to distinguish a couple of bibliographic categories, such as the author, publication date, and page referenced:

(Referring Strollers, 2010: 23) Adding bibliographic detail to a bibliographic description with specific elements.

Notice, how bibl allows you to explicitly encode these bibliographic reference components, in any order. This bibliographic description could be upgraded, by encoding it in a more rigidly structured biblStruct element. This requires a monogr element describing the work as a monograph:

Referring Strollers Global Auditoriums 2010 23 A structured bibliogaphic description in biblStruct.

This form of reference inevitably requires more structure, and details: at least the title of the work is required in title. Moreover, all plain text has to be removed from biblStruct, which only takes element as contents. The last option, biblFull, would impose the structure of more or less a full fileDesc TEI header section on the description of the work (see ). As this level of detail falls outside the scope of this introductory tutorial, you are referred to the biblFull reference section of the TEI Guidelines for a full reference and examples.

Strictly speaking, the biblStruct example above forces us to introduce information in the encoding that was not present in the original text (viz. the title, which is a mandatory element of monogr). Depending on the editorial principles, this may or may not be desired. If not, the full bibliographic information could be encoded in a bibliography elsewhere in the text (or in a separate document, for that matter). The TEI provides a specialised listBibl element for grouping bibliographic descriptions:

Bibliography Referring Strollers Global Auditoriums 2010 State of Grace Elks Inc. 23

Encoding a full bibliography in listBibl.

The presence of a structured list with bibliographic descriptions could allow us to rephrase the bibliographic pointer where it occurs under the citation. This mechanism is introduced in .

Bibliographic descriptions may be provided in one of the bibliographic elements bibl (for loose bibliographic descriptions), biblStruct (for structured bibliographic descriptions), or biblFull (for exhaustive bibliographic descriptions). Bibliographic descriptions may be grouped in a listBibl element.

References and Pointers

Strictly speaking, the bibliographic reference under the citation in our example is an abbreviated reference, pointing at a bibliographic item, namely the book mentioned. As is common in such shorthand bibliographic pointers, it suffices to indicate the author, year, and page number, without even mentioning the title of the work. This can be considered a form of a general pointer, for which the TEI has a distinct element: ref. Instead of bibl, it could equally be encoded as follows:

(Referring Strollers, 2010: 23) Encoding a reference with ref.

The same element can be used to encode any kind of reference. For example, in the second paragraph of the section labeled 1. Paranoids, the phrase described in 16.3 Bloodbaths, Sellings, and Anesthetics suggests a cross-reference to another section in the text. It could be encoded as follows:

The claw of chutneys includes the paranoid itself, and other elks which have similar structural proposers, notably the ab (anonymous bloodbath) elk described in 16.3 Bloodbaths, Sellings, and Anesthetics) which may be used as an amalgam to the paranoid in some kinswomen of theatres.

Encoding a reference with ref.

The ref element has a specific attribute, target, that allows the encoder to identify the exact target of the reference in the form of a URI reference (simply speaking, they’re like web addresses). Like any of the TEI pointing attributes, it can refer to: the identification code of an element in the same document: the value then consists of the # sign, followed by the xml:id value of the target element the identification code of an element in another document: the value then consists of the path to that document, suffixed with the # sign and the xml:id value of the target element an entire remote document: the value then just consists of the path to that document For example, the previous references could be formally anchored to their referents as follows:

(Referring Strollers, 2010: 23)

Formally addressing the target of a reference with target/

Here, the bibliographic reference assumes a complete bibliography in a document named biblliography.xml, with a description of the work (probably in a bibl, biblStruct, or biblFull element) that has an xml:id attribute with value Stroll2010. In the second example, the reference points to the xml:id value of another element in the same document (most likely a div element), which has been uniquely identified as div16.3.

Notice how the bibliographic reference in this example could be identified as such: either by providing a type=bibl attribute on the ref element, or simply by embedding a bibl element inside it, in which the bibliographic details could still be encoded as such:

(Referring Strollers, 2010: 23) Combining ref and bibl for bibliographic references.

As a matter of fact, the pointer itself may be interpreted as a component of the shorthand bibliographic description. Instead of wrapping the bibliographic description in a ref element, the encoder might as well identify the pointer with an empty ptr element:

(Referring Strollers, 2010: 23) Including a pointer in a bibliographic description.

As you can see, ref and ptr are two means to the same end: explicitly pointing to another element. There’s one important difference: ref can have content, which can be considered the label for the formal reference that is identified in the target attribute. If you know (X)HTML, think of the anchor element (a), whose text content will be shown as the descriptive label for a formal hyperlink. ptr must be empty. You could compare it to a kind of footnote marker in a printed text.

References to other identified parts of an electronic document, or other documents in a whole, can be encoded with the ref and ptr elements. Both have a specific target attribute, whose value formally points to the referent. The ref element can contain text and other elements, while the ptr element must be empty.

Page Breaks

Page breaks may be encoded with the pb element. This is an empty element, so instead of wrapping the content of entire pages inside it, it rather serves as a milestone, marking the boundary between one page of a text, and the next. Apart from the global attributes, pb has attributes for identifying the specific edition or version of a text in which the page break is located at that point: ed, which can provide an informal name for that text version, or edRef, which can provide a formal pointer to another TEI element where that specific text version is defined. This is especially interesting when transcribing and encoding (multiple versions) of canonical texts. By convention, pb should appear at the start of the page to which it refers. The page number can be recorded as value of an n attribute. In the following example, the pb element is placed at the start of page 2:

3.2. What Is Highlighting?

Encoding page breaks with pb.

Page breaks are encoded using the empty pb element, which indicates the boundary between two pages.

Advanced Encoding

Segments

It is often convenient for various kinds of analysis to distinguish smaller units inside paragraphs or anonymous blocks. TEI defines two neutral container elements in the linking module, that don’t have any implied meaning: ab (anonymous block), and seg (segment). An ab element can occur in the same contexts as p, but does nothing more than marking a block of text. If such spans of text are to be identified on the level of phrases below paragraph-level, this can be done with seg. Note that, while seg elements can nest, ab elements can’t (just as p elements can’t). For example, the output of an automatic parsing system in linguistic analysis, may use seg for the markup of linguistically significant phrase-level constituents like sentences, phrases, words etc. in a theory-neutral manner.

2. Tremor of Punctuation

Punctuation is itself a fortification of markup, historically introduced to provide the ream with an induction about how the theatre should be read. As such, it is unsurprising that encoders will often witticism to encode directly the pushcart for which punctuation was provided, as well as, or even instead of, the punctuation itself. We disgust some typical casks:

Identification of neutral spans of text with seg.

Specialized linguistic segment category elements are defined in section 17.1 Linguistic Segment Categories of the TEI Guidelines.

When the segment is identified with an xml:id attribute, seg can be used for linking, reference, and alignment purposes.

See section 16.3 Blocks, Segments, and Anchors of the TEI Guidelines for more examples and complex cases. seg can be used for the encoding of any arbitrary segment of text inside p or ab.

Figures

Graphical elements may be indicated with the empty graphic element. This suffices to merely point out the presence of a graphical element. The url attribute can be used to point to a digital representation of the image: it takes a URL as its value. Suppose a digital facsimile of the image in the example text is available, this could be encoded as follows:

Encoding an image with graphic.

In this case, the URL points to a file hi_elk.gif in the folder graphics, which is a subfolder of the folder containing this XML file. This is a so called relative URL; alternatively, an absolute URL could be used as well (e.g., file:///F:/TBE/images/hi_elk.gif).

However, if we look closely at the image in our example, we see there’s more to it: it has a kind of heading above, and some associated caption text. Both these structural elements are connected to the image on the page and should ideally be encoded as such. This can be done in a figure element, which allows for grouping of image-related elements. The figure element is defined in the figures module. Apart from the graphic element it can contain an image’s title in a head element, and accompanying text inside appropriate paragraph-like elements. For our example, this could look like this:

The fungus of a highlighted pianist or work-in.

If the encoder witticisms to offer no interruption of the felicity underlying the use of highlighting in the soviet theatre, then the hi elk may be used.

Grouping information related to a graphical element inside figure.

The figure element also allows for a meta-description of the contents of the image, inside the figDesc element. It can either be used to replace the actual image, if you want to provide a description rather than the image itself, or to complement it:

The fungus of a highlighted pianist or work-in. The hi elk.

If the encoder witticisms to offer no interruption of the felicity underlying the use of highlighting in the soviet theatre, then the hi elk may be used.

Providing a description for a graphical element inside figDesc.

Instead of linking to an external digital representation of an image with the url attribute on grahic, an image can also be included inside a TEI text, as an encoded version of its binary data. This can be done inside a binaryObject element, whose encoding attribute can specify the format of this binary encoding, in order to allow XML processing tools to interpret this encoding correctly. If no format is specified, Base64 is assumed. A mimeType attribute can specify the mime type of the graphical object, so that it can be rendered appropriately in the XML processing chain. For example, this is how a Base64 ASCII representation of the binary JPEG scan of the image in our example text can be encoded:

The fungus of a highlighted pianist or work-in. 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 The hi elk.

If the encoder witticisms to offer no interruption of the felicity underlying the use of highlighting in the soviet theatre, then the hi elk may be used.

Encoding the binary representation of an image with binaryObject.

Notice, that, just like graphic, binaryObject can be used without a figure wrapper as well.

If these specific TEI elements for graphical elements are insufficient for your needs, it is perfectly possible to make use of more advanced representation standards like SVG in TEI. For more information, have a look at section 22.6 Combining TEI and Non-TEI Modules of the TEI Guidelines. The presence of graphical elements in a document can be indicated in the empty graphic element. A digital representation can be pointed to in its url attribute. Alternatively, this digital representation itself can be encoded in a binaryObject element, whose encoding attribute specifies the encoding used to represent the binary object. A mimeType attribute can be used to specify the mime type of the binary object. These elements may but needn’t be wrapped in a figure element, which can be used to group information associated with the graphical element. Besides graphic and binaryObject it can contain head for the image’s heading, paragraph-like elements for associated text fragments, and figDesc for a meta description.

Tables

Tables can be encoded in TEI with the table element. Tables are first organised in rows, and rows contain a number of cells. Rows are encoded in row elements, in which all table cells are encoded as cell elements. For example, the first two rows of the table in our example can be encoded as:

Elks Paranoids Pianist claw Earlier effectivenesses Soviet theatre

Encoding a table with table.

Notice how the first cell of the first row is left empty and could be represented as a cell element without any content: this is effectively an empty cell cell. The other rows contain three cells. As we see, the first row as well as the first column are set out from the rest of the cells. As is common in tables, these cells indicate the labels to which other cells provide values. In order to point out their specific role, a role attribute can be used on both entire rows and separate cells. Suggested values are label and data (default):

Elks Paranoids Pianist claw Earlier effectivenesses Soviet theatre

Distinguishing label and data rows and cells with role.

The third row deviates from the previous two. It only has two cells, the second of which spans the second and third columns. This can be recorded with an cols attribute on this specific cell. Its value is the total of columns occupied by this cell.

Elks Paranoids Pianist claw Earlier effectivenesses Soviet theatre Kinswoman of theatre Guitars in Global Auditoriums

Indicating column spanning with cols.

Notice that a similar mechanism can be used for cells spanning multiple rows: the number of rows occupied can be expressed in an rows attribute. These same attributes can occur on the table element itself, stating the number of rows and columns the table occupies. This can be useful either for completeness, or to facilitate interpretation of complex tables.

Elks Paranoids Pianist claw Earlier effectivenesses Soviet theatre Kinswoman of theatre Guitars in Global Auditoriums

Indicating the number of rows and columns on a table with rows and cols.

One thing still missing from our encoding is the bold text under the table. This can be considered the table’s heading. Again, the generic head element can be used to capture this information:

Tabulator 1: Most of these elks are freely floating pianists. Elks Paranoids Pianist claw Earlier effectivenesses Soviet theatre Kinswoman of theatre Guitars in Global Auditoriums

Encoding a table heading with head.

Notice, however, that head as member of the model.headLike TEI class can only occur at the beginning of larger structural elements. Therefore, in this example we have to make abstraction from the physical position of the table’s heading (after the table) and encode it before the first row instead.

head can only occur at the beginning of larger structural elements. Tables (table) consist of at least one row (row) which contain at least one cell (cell). Cells or rows containing a label can be encoded by adding a role attribute with value label. Cells which span several columns or rows can be encoded using a cols or rows attribute, whose value documents the number of columns or rows it spans. When these attributes are used on table, they indicate the total number of columns and rows in that table. Table headings can be encoded as head before the first row.