Are characters divided into meaning part and phonetic part, with no overlapping?

In short, yes, the phonetic part may or may not contribute to meaning. But let me put this more precise. I understand that Western readers must have been bewildered by the strange history of Chinese characters. There has been so much misconceptions that I cannot resist to add several remarks.

Abbreviations:

char := Chinese character
p(*) := pronunciation of *
m(*) := meaning of *
x ~ y := x is close or related to y
x ~/~ y := x is not apparently close or related to y

Consider char AB consisting of char A and char B sticking together (vertically or horizontally, or stuck in the middle). For our purpose, let char AB be an 形聲 (xíng shēng / Phono-semantic character) char. Without loss of generosity, let char A be the phonetic part, and char B be the semantic part. By our assumption,

p(AB) ~ p(A) ... ... (a)
m(AB) ~ m(B) ... ... (b)

Logically, in our setting, there are these possibilities: (Examples follow shortly)

m(AB) ~ m(A) ... ... (c)
m(AB) ~/~ m(A) ... ... (d)

Before seeing some examples, we naturally wonder, why did this happen? How was the accepted form of AB be agreed on, in the Chinese-using community? And how did these consensus arrive among them at all? You might picture that (along evolution of time):

• Suppose p(A) = P and m(B) = M
• It was intended that a new char X be created with m(X) = M' and p(X) = P', with P' ~ P and M ~ M', but either it happens that m(A) ~ M' (case c) or m(A) ~/~ M' (case d).
• Then elders of the tribe or government officials (??) gathered, to decided how X is written.
• They decided X = AB.
• They announced the decision, and citizens (??) happily accepted.

The process is not only strange-sounding but in fact, wrong. There are two ways how it happened.

---

The first version is 假借 (borrowed homophone). This corresponds to case d above.

• A concept M' with sound P' had already been established in spoken Chinese, but no corresponding char existed.
• Someone (unlikely a she in the patriarchal society) wanted to write down a char X with m(X) = M' and p(X) = P'.
• He made great effort to find a char A with P = p(A) ~ P', but N = m(A) ~/~ M' in general.
• He wrote A, intending m(A) = M' and p(A) =P'.
• From now on, m(A) = N or m(A) = M', depending on context. A is thus overloaded.
• Since so much confusion eventually arose, people started to add another char B with M = m(B) ~ M', that is to say, to coin AB, replacing the second meaning of A. (But Q = p(B) ~/~ P' in general.)
• Thus, p(AB) = P' and m(AB) = M'.

Examples:

莫(Oracle), a picture of setting sun among woods, means "evening". It was borrowed to be the homophone 莫 "not", so 日 "sun" (a circle with a dot showing the Sun) was concatenated to it, becoming "暮".

父(Oracle), a picture of a sharp tool, means an ax. It was borrowed to be the homophone 父 "father", so 斤 "a small ax" (a picture of a sharp tool) was concatenated to it, become "斧".

---

Another version is 引申 (figurative meaning). This corresponds to case c above.

• A new concept M' was gradually formed, calling for written expression yet to exist.
• There was the first person who wished to write some X with m(X) = M'.
• He made great effort to find a char A with N = m(A) ~ M'.
• He wrote A, intending m(A) = M', while p(A) =P, as was before.
• Since so much confusion eventually arise, people started to add another char B with M = m(B) ~ M', and to coin AB, replacing the second meaning of A. (But Q = p(B) ~/~ P' in general.)
• Phonology has changed so much that p(AB) = P' ~/~ P = p(A), no longer same, and people do not remember how it was created.

Examples:

冓(Oracle), a picture of a pile of crossed wood sticks, means "to connect, to join". It gave rise to figurative meaning "to talk", so 言 "speech" (a picture of a woodwind instrument) was concatenated to it, becoming "講".

卑(Oracle), a person standing, holding a tablet or plaque, means probably a servant job (modern sense: inferior, bad). It gave rise to a related and more peculiar sense "maidservant", so 女 "woman" (a picture of a dancing person) was concatenated to it, becoming "婢".

---

It hardly (if ever) was the case that X = AB came into existence without the intermediate stage of borrowed homophone or figurative meaning. Everything becomes reasonable, doesn't it!

---

ref.: Some of them comes from my memory, sorry. Some books I once borrowed or read are not at hand, but I cite one book I have, 許進雄《簡明中國文字學》(台灣新北市:學海出版社,2000).

Take for example classical character 講 = 言(talk) + 冓 (a strucutre/scaffold from bamboo twigs).

So, the meaning is "talk+structure", thus a structured talk which is "a lecture". You see that 講 doen't have a phonetic information since 冓 sounds as "gou". It is written today as 讲, which is 言+井(A square well).

井 was added as a phonetics (it is spelled as "jing"), but it is not only a phonetic component. It is also a simplification of scaffold image

so it has nothing to do with a meaning of "a well". You can see that 井 still looks like a scaffold/structure.