凡是含有聲符的字統稱為「有聲字」。「有聲字」可以把甲骨文中含有聲符的字全部納入其中。本文由兩部分組成。在第一部分「有聲字的構造類別」中，我們認為可分為三個層級十種構造類別：（一）獨體形聲字；（二）附劃因聲指事字；（三）兩聲字；（四）「从某，某聲」形聲字；（五）亦聲字；（六）省形字；（七）省聲字；（八）既省形又省聲字；（九）多形字；（十）多聲字。這十種類別並非處於同一個分類層級上。第一層級是「有聲字」。第二層級是（一）至（四）四種類別，隸屬於「有聲字」。第三層級是（五）至（十）六種類別，隸屬於形聲字，為形聲字的變體。跟小篆相比，可以知道，（一）至（三）為《說文》小篆所淘汰，而（四）至（十）被繼承下來。由此說明，漢字的構造類別因時代的不同而不斷更新。在第二部分「形聲字産生的途徑」中，我們認為有九種産生途徑：（一）形、聲相配；（二）形、聲裂變；（三）變形聲化；（四）加注形符；（五）加注聲符；（六）形符代換；（七）聲符代換；（八）形符的繁與簡；（九）聲符的繁與簡。其中，加注形符和聲符是形聲字産生的主要途徑。（六）至（九）是形聲字産生的途徑，也是形聲字異體字産生的途徑。

The term “phonetic characters” refers to all characters that have a phonetic symbol, and may include those characters with phonetic symbols in the oracle-bone inscriptions. This paper is composed of two sections. In the first section, “The Structural Classifications of Phonetic Characters,” we sort phonetic characters into ten structural classifications broken into three different classes. The ten structural classifications are as follows: (1) single-component pictophonetic characters, (2) self-explanatory characters enhanced by a single-component pictophonetic character with a special added symbol, (3) bi-phonetic characters, (4) pictophonetic characters in which one element denotes meaning and the other sound, (5) pictophonetic characters in which the phonetic symbol indicates both meaning and sound, (6) pictophonetic characters in which several strokes are omitted from the pictographic element, (7) pictophonetic characters in which several strokes are omitted from the phonetic element, (8) pictophonetic characters in which several strokes are omitted from both the pictographic and phonetic elements, (9) pictophonetic characters with multiple pictographic elements, and (10) pictophonetic characters with multiple phonetic elements. These ten structural classifications do not belong to a single class. “Phonetic characters” make up the first class, while the structural classifications (1) – (4) belong to the second class and are subsidiary to “phonetic characters.” The structural classifications (5) – (10) are placed in the third class and are subsidiary variants of pictophonetic characters. When compared with the Lesser Seal script, it is clear that structural classifications (1) – (3) were selectively excluded from the “Shuo Wen Jie Zi” 說文解字, the first Chinese dictionary compiled by Xu Shen 許慎 in 121 A.D., but structural classifications (4) – (10) remained included. From this it is evident that Chinese characters were constantly updated as times changed. In the second part, “The Methods by which Pictophonetic Characters are Produced,” we conclude that there are nine methods by which these characters are formed: (1) the pictographic element and the phonetic element match, (2) the pictographic and phonetic elements split into two, (3) part of the pictographic element in certain pictographic characters, associative compounds, and self-explanatory characters becomes a phonetic element, (4) a pictographic symbol is added to produce a pictophonetic character, (5) a phonetic symbol is added to produce a pictophonetic character, (6) the pictographic elements are substituted, (7) the phonetic elements are substituted, (8) the simplicity and complexity of the pictographic element, and finally (9) the simplicity and complexity of the phonetic element. Among these, the primary methods by which a pictophonetic character is produced are by the addition of a pictographic or phonetic symbol to an existing character. While pictophonetic characters may be produced via methods (6) through (9), these methods may also produce variant forms of pictophonetic characters.