Types of written characters

Introduction

From the 17th to the 18th centuries, Western scholars considered Chinese characters and Egyptian hieroglyphs to be ideograms (symbols that convey only meaning), i.e., the opposite of phonograms (symbols that convey only sound). This misconception is still quite common among non-specialists.

In academic circles, such a dichotomy has long been abandoned, and a triple system is still often used today: phonograms – ideograms – logograms. Chinese characters are classified according to this system as logograms (symbols that convey words – i.e., both sound and meaning).

However, this system has one significant problem – it is designed to describe a writing system as a whole, while all early writing or proto-writing systems (e.g., Egyptian, Sumerian, Anatolian, Mayan, Dongba) contain symbols of all three types (and modern ones usually two – phonograms and logograms), and therefore these writing systems as a whole cannot be classified as belonging to any one specific type.

This article will present a modern system that divides all written characters according to clear criteria suitable for any writing system (with an emphasis on Chinese and Egyptian, since these are the early writing systems I am most familiar with).

Phonograms – Ideograms – Logograms

First, let’s consider this common system using examples of symbols from some early writing systems. It is based on two criteria:

Logograms: The Egyptian hieroglyph ⟨𓉐⟩ represents the Egyptian word {pr “house”}; the cuneiform sign ⟨𒆳⟩ represents the Sumerian word {kur “mountain”} or the Akkadian word {šadûm “mountain”}.

Phonograms: The Egyptian hieroglyph ⟨𓅓⟩ represents the sound [m]; the cuneiform sign ⟨𒈬⟩ represents the sound [mu]; the Anatolian hieroglyph ⟨𔗷⟩ represents the sound [a].

Ideograms or Determinatives: The Egyptian hieroglyph ⟨𓂻⟩ conveys the idea of “TO GO, TO MOVE”; the cuneiform sign ⟨𒄑⟩ conveys the idea of “TREE, WOODEN”; the Anatolian hieroglyph ⟨𔖖⟩ conveys the idea of “DEITY”.

It is worth noting that these types are not fixed for specific symbols but can easily change depending on the context. For example, compare the use of the Egyptian hieroglyph ⟨𓆱⟩:

Groups of symbols

Before moving on to the main part, it is necessary to talk about an important thing – groups of symbols. In ancient writing systems, several symbols could be combined into groups.

The picture above shows an example of the grouping of Egyptian hieroglyphs – the hieroglyph ⟨𓆑⟩ is located under the hieroglyph ⟨𓌬⟩, and “inside” the hieroglyph ⟨𓆓⟩ are ⟨𓏏⟩ and ⟨𓇿⟩. In Egyptian script, grouping is done only for the visual beauty of the text, but in the following example, it is meaningful.

In Mayan script, grouping plays a very important role – the word often acts as the “basis” of a block, and various auxiliary phonetic or semantic symbols are added at the edges. In the example above, the bases are the logogram {CHUM} “to sit”, to which the phonograms [la] and [ja] are added; {AJAW} “king”, to which [ti], [wa] and [le] are added; and {K’UK’} (name), to which the logogram {AJAW} “king” is added. The whole sentence will look like this (the sentence is glossed quite differently from how it is written, but I will not give the glosses here):

CHUM-la-ja ti-AJAW-wa-le K’UK’-AJAW*
“King K’uk’ sits down on the throne”*

Let’s look at an example from another writing system.

The image above shows an inscription in the Anatolian script from one stele. Written in the Luwian language, it says the following:

| wa/i-mu | á-mi-zi-i | tá-ti-zi DEUS-ni-zi-i | (LITUUS)á-za-ta*
“The gods of my ancestors loved me”*

Here, grouping plays not only a decorative role, but also makes the text more understandable to the reader – the symbols are divided into groups in such a way that the structure of the sentence is better analyzed.

The last example can be compared to the modern use of spaces – and here we move on to defining a group not so much as a graphic device, but as a way of combining several symbols to write one meaningful unit of spoken language (usually a word).

Thus, any word written, say, in English, Russian, and other languages with alphabets, can also be considered nothing more than groups of symbols. For example, the group ⟨kid⟩ consists of three phonograms ⟨k⟩, ⟨i⟩, and ⟨d⟩.

Let’s also look at the familiar division of Chinese characters, where each character = one logogram representing one word.

Although parts of Chinese characters are usually considered “elements”, they should rather be regarded as independent symbols, and therefore, in fact, compound Chinese characters are nothing more than groups of symbols.

For convenience, characters consisting of one symbol should also be understood as groups, since in the context of Chinese script, a group = a character, and it consists of graphemes, i.e., in groups consisting of a single grapheme: grapheme = character = group.

Unlike Egyptian or Sumerian, the words of the Old Chinese language were mostly monosyllabic, so grouping was easy and extremely rapid, which eventually led to a huge number of different “characters” perceived as a single unit.

There are also different layers of character grouping. Take, for example, Japanese script.

Here the word {kanji “Chinese character”} is written using the symbols ⟨漢字⟩, the reading of which is written above in phonetic symbols as ⟨かんじ⟩. This entire combination can be considered one large group representing one word. In this group, however, it is possible to distinguish subgroups, and in them even smaller elements can be distinguished (how small – is decided by whether we want to give a synchronic or diachronic description). Below is a hierarchy of groups (highlighted in red) with a description according to the modern systematics (see below):

Ligatures

Ligatures are essentially established groups of symbols that can be considered independent symbols. For example, above was given the Egyptian hieroglyph ⟨𓌬⟩, which is a ligature of ⟨𓌫⟩ and ⟨𓂻⟩. The cuneiform glyph ⟨𒈗⟩ is a ligature of ⟨𒇽⟩ and ⟨𒃲⟩, and the Anatolian glyph ⟨𔑾⟩ consists of ⟨𔑺⟩ and ⟨𔖻⟩. In essence, developed systems, such as Demotic or modern Chinese script, are almost entirely composed of ligatures that are treated as independent symbols rather than groups.

Modern Systematics of Written Characters

Now let’s move on to modern theory. It was developed by Egyptologists, but surprisingly enough, it is suitable for describing any writing system in general. It is based not on two, but on three criteria:

The main source for this classification is Stéphane Polis, Serge Rosmorduc, 2015. The Hieroglyphic Sign Functions: Suggestions for a Revised Taxonomy.

Ideograms

Ideograms are symbols that do not carry any sound and are not tied to any words; their purpose is to convey only meaning. Usually ideograms are pictograms – symbols that directly depict an object or a certain plot. For example, a machine operating manual consists entirely of ideograms.

Let’s also look at one example from Egyptian script:

mḥ.j pr-Rꜥ m špsw-ꜥnḫ qn.w m tw.wt ḥr X ḥr Y
“I filled the temple of Ra with many sphinxes and statues like X and Y”

Here, images of statues are used simply as a demonstration of what kind of statues were placed in the temple – these symbols do not convey any words, they are just pictures.

Another good example is the Dongba proto-writing, which itself is largely composed of ideograms.

The egg was thrown into the lake. A white wind blew from the left, and a black wind blew from the right. Because of the winds, ripples appeared on the lake. The waters of the lake threw up the egg, it collided with a cliff, and a radiant creature was born from it.

Actually, the language is not fully transmitted here, thus it is still a proto-writing. The only non-ideographic symbols here are (1) the phonogram for the word “white” (above the “wind on the left”), derived from the logogram for the word “to unite”; (2) the logogram for the word “black” (above the “wind on the right”); (3) the logogram for the word “cliff”, supplemented by (4) a phonetic, derived from the logogram for the word “chicken”, depicting a chicken head.

Logograms

Logograms are symbols that directly convey words, with their sound and meaning. Some of them represent a specific object:

They can also convey a word through an object associated with it:

In Egyptian, the word {nḏm “sweet”} is conveyed by the image of carob beans, the word {ṯꜣw “wind”} by the image of a sail, and {jrṯ.t “milk”} by a bottle of milk.

In Anatolian, the word {hantawatt- “king”} is conveyed by the image of a king’s hat, the word {wasu- “good”} by a triangle (symbol of happiness), and {tuppala- “secretary”} by a stone tablet with a chisel.

In Chinese, the word {ɢʷaŋ “king”} is conveyed by the image of an axe (symbol of the king’s power), the word {lˤats “big”} by the image of an adult, and {kˤau “high”} by the image of a tall building.

Logograms can also be polyvalent, i.e., they can convey several different words at once.

Phonograms

Phonograms are symbols that convey only sound, without regard to the meaning. They usually develop from logograms to write phonologically similar but semantically different words (this method of writing is often called “rebus” or, in the context of Chinese script, “phonetic borrowing”).

For example, the Egyptian logogram ⟨𓆱⟩ {ḫt “tree”} can act as a regular phonogram with the reading [ḫt] (see the example above). Phonograms do not always take the reading of logograms entirely, for example, the Anatolian logogram ⟨𔕮⟩ {sasanza “seal”} writes [sa] as a phonogram, and ⟨𔑇⟩ {piya- “to extend (hand)”} writes [pi]; the Egyptian logogram ⟨𓈎⟩ {qꜣꜣ “hill”} writes [q], and ⟨𓉐⟩ {pr */ˈpaːɾuw/ “house”} writes [pr].

Also, they do not always exactly copy the reading of the original logogram, but can be used for phonetically similar variants. For example, the Sumerian logogram ⟨𒀭⟩ {an “heaven”} can write [an] and [am], and ⟨𒈬⟩ {mu “name”} can write [mu] and [ŋu]. This is clearly seen in the example of Chinese script – for example, the logogram ⟨石⟩ {dak “stone”} can phonetically write words like {tak “to pick up”}, {tˤaks “to be jealous”}, {staks “numerous”}, {tʰˤak “bag”}, {dˤaks “measure”}, etc., that is, as a phonogram it writes some general sound sequence [TAK].

Many characters can be homophonous, so different characters can be used to write the same combination of sounds – often interchangeably, but also according to established written tradition. For example, [du] in Sumerian can be written in more than ten different ways, but in the words {du “to hit”}, {šudu “completely”} and {hedu “decoration”} the syllable [du] is usually written as ⟨𒌌⟩, and in {du “to build”}, {šudu “to reproach”}, and {gidu “fence”} it is usually written as ⟨𒆕⟩. The Old Chinese particle {ʔan} in the Warring States period was written as ⟨焉⟩ in the western and as ⟨安⟩ in the eastern states – this is nothing more than an orthographic habit.

Because of the polyvalence of logograms, phonograms can also convey several unrelated readings. This can be clearly seen in the Chinese script, where, e.g., the logogram ⟨勺⟩, conveying two words {mtewk “ladle”} and {pew “handle of a ladle”} (see above), can convey two different phonetic qualities: [PEWK] for words like {nəprˤewks “leopard”} or {brˤewk “small melon”}, and [TEWK] for words like {tewk “to pour wine”}, {tˤewk “reins”} or {dewk “spring”}.

It is sometimes quite difficult to draw boundaries between logograms and phonograms, because a phonetically borrowed meaning can become the main one, and the character will acquire the reading and meaning of a completely different word, losing the original one. For example, the logogram ⟨𓅭⟩ {zꜣt “duck”} was phonetically borrowed to write the word {zꜣ “son”}, and over time the latter word became the main one for the hieroglyph, i.e. ⟨𓅭⟩ for {zꜣ “son”} moved from phonogram to logogram.

Radicograms

Radicograms are essentially non-independent logograms. They are used as phonetic and semantic complements to etymologically related or semantically similar words. For example, the Egyptian hieroglyphs ⟨𓄚⟩ and ⟨𓂙⟩ can both convey the sound [ẖn], but ⟨𓄚⟩ can be used in any word (i.e., as a regular phonogram), for example, to write the words {ẖn.w “inside”} or {ẖn “to approach”}, whereas the use of ⟨𓂙⟩ is mainly limited to the word {ẖnj “to row”} and those etymologically related to it, such as {ẖnw “rower”}, {ẖn.t “to sail in a boat”}, {mẖn.t “transport ship”}, etc. (and even occurs once to write the word {ẖntj “statue”}, which in context was specifically a statue used in water processions) – i.e., it conveys not only sound, but also meaning.

In Chinese, it is also possible to distinguish a certain number of radicograms within characters, for example, ⟨四⟩ {slis “four”} within the character ⟨駟⟩ {slis “team of four horses”} can be considered a radicogram.

The boundaries between radicograms and phonetics are even more blurred than between logograms and phonograms, so it is quite difficult to distinguish them.

Semantics

Semantics or determinatives are auxiliary symbols designed to hint at the meaning of a particular word written by a group of characters. Compare the Egyptian words in the picture:

In all cases, the same groups are used, conveying the reading [wn], but the semantics are different in all cases: ⟨𓉿⟩ “DOOR”, ⟨𓂻⟩ “MOVEMENT” and ⟨𓅪⟩ “BAD”.

Semantics can also be added to logograms (especially polyvalent ones), specifying their meaning:

Chinese script is no exception. For example, the character ⟨叚⟩ had a phonetic value of [KA] and as a phonogram originally recorded words such as {krˤaʔ “to borrow”}, {krˤa “intestinal obstruction”}, {ɡrˤas “free time”}. Later, semantics were added to it:

Phonetics

Phonetics are auxiliary symbols designed to hint at the reading of a word written by a group of characters. They can be added to both logograms and phonograms.

In the case of Chinese script, it is not always easy to say which character is dependent and which is independent and the original one. It is also not always possible to say whether the dependent character was added to the original at all, or whether the group of characters was assembled from scratch from two separate characters.

Phonetics can also form ligatures with a character, sometimes replacing an existing element in it.

Diacritics

Diacritics or decorative/differentiative marks are auxiliary symbols that carry neither semantic nor sound function, but are adopted either to graphically decorate a symbol or to differentiate it by reading or meaning.

In modern scripts, diacritics are quite common – compare, for example, the letters of the Russian Cyrillic alphabet ⟨е⟩ [ʲe] and ⟨ё⟩ [ʲo] or the syllables of the Japanese hiragana ⟨か⟩ [ka] and ⟨が⟩ [ga].

In Chinese script, decorations can be used for simple graphic complication of characters:

They were also used for semantic differentiation of polyvalent logograms:

Another use is to indicate the phonetic usage of a particular logogram:

In Egyptian, diacritics can be used in the same way as in Chinese – to differentiate polyvalent logograms:

Also in Egyptian, small dashes can be added to simply distinguish between different but graphically similar hieroglyphs. For example, one of the forms of the hieroglyph ⟨𓏒⟩ is ⟨𓏔⟩, but since there is a completely different homographic hieroglyph ⟨𓏔⟩, dashes ⟨𓏕⟩ can be added to that form to clearly indicate that this is the form ⟨𓏒⟩, and not ⟨𓏔⟩.

Punctuation Marks

Punctuation marks are independent symbols that convey neither sound nor meaning, serving only to separate some elements of the text. This type will not be described here in detail, because, i think, it’s already clear – these are symbols like a period ⟨.⟩, a comma ⟨,⟩, a hyphen ⟨-⟩, brackets ⟨( )⟩, etc.

Egyptian script also has special cartouches that work roughly like parentheses, and which enclose various important names: