Japanese Linguistics #1: The four major categories of kanji characters

A friend lent me her copy of her textbook Nakama 2: Japanese Communication, Culture,Context a short while back so that I could do some Japanese self-study in the holidays. Here's something I learnt after reading "Types of kanji 5: Historical perspective", which I thought others would see as a useful foundation to learning kanji - the shared script of both Chinese and Japanese.

Pictographs

These characters are derived from physical images and objects. There is a distinguishable link between their shape and the object they are describing. They are among the first kanji characters that a beginner Japanese learner would be acquainted with.

Examples include 日 (sun) and 火 (fire).

Logographs

Some concepts are abstract and difficult to depict pictographically. This category of kanji represent abstract ideas rather than a physical object.

For example, the numbers 一, 二 and 三 need to explanation. The horizontal line in 上 and 下 represent the horizon, with lines extending above and below the horizon respectively to convey direction.

Semantic compounds

These characters are comprised of two or more simpler kanji characters. Combined, they convey more diverse expressions.

For example, the character 明 (bright) is made up of 日 (sun) and 月 (moon). The character 林 (woods) is made up of two 木 (wood) characters, whilst 森 (forest) has three. Go figure.

While we're on this topic, one character I've admittedly always been puzzled by is 暗 (dark). Why is it made up of two 日 (sun) and a 立 (upright)…? Anyone got any input?

Phonetic compounds

This is by far the largest group, constituting 80% of kanji. Which sucks, because their meanings are so easy to mix up -.- They are typically made up of two components, whereby one component indicates its semantic categorisation, whilst the other component indicates the on-reading. This might sound a bit confusing, so here's some examples.

The characters 晴, 静 and 靖 both have the character 青 in common, and they all share the same on-reading (セイ). This is because the component they have in common is the phonetic component.

However, the characters 緒, 級, 納, and 続 which all have the component 糸 in common do not have a similar sounding on-reading at all. This is because the comopnent in common is the semantic component.

How do you tell if the component you recognise is the phonetic or the semantic component? All I've been able to do is to construct a table of similar looking kanji and remember each one. The only other suggestion I have is to read widely and learn each character in its context. Anybody have an easier proven method to distinguish different phonetic compounds?