The Possibility of Computational Art and Culture (6) - Hideo Kobayashi's Stance - Similar to the process of 'continuing to calculate with high precision without truncating floating-point numbers (fine numerical values)'

Contradictions and conflicts are often themes in art, and as Jaspers also says, it is essential for us not to reach easy answers, but to embrace contradictions and continue thinking.

When we think through things thoroughly, we almost inevitably face contradictions, and perhaps thinking that does not encounter contradictions is merely half-baked thinking.

The writer Hideo Kobayashi seems to teach us that a work of art can only be created after thinking through things so thoroughly.

Hideo Kobayashi seems to have viewed contradiction as something positive.

Regarding this, in 'On History,' Hideo Kobayashi says,

“Even if you call it scholarship, Norinaga's scholarship is a study of life.
It is not a study of matter.
It is fine to find the interest of contradiction there.
Just as a novelist explores the interest of contradiction in life.
The fact that what a person says must be contradictory can also mean that the person was thinking that deeply.
If you stop thinking a little earlier, nothing would have to be contradictory; such things also exist.
You can think that there is a structure in the brain where contradictions arise when you think things through.
Norinaga did it knowing what he was doing.
He didn't contradict himself because he was a fool.
He expressed contradictions in a very clear and explicit form, but this is something that exists in any true thinker.”
By saying this, it is likely that Hideo Kobayashi was sensitive to the problem of contradiction in literature—an art form where contradictions and conflicts are themes—because he was involved in literature and was constantly thinking about the concept of contradiction.

In literature, because the fact that humans are beings burdened with karma and contradictions, as well as human folly and sorrow, is depicted, perhaps literary figures like Hideo Kobayashi did not assume they carried justice or preach justice.

It seems to me that not fearing contradiction also means not falling into simple dualism of good and evil.

Just because one side is good does not necessarily mean the other is evil; there can be good and evil simultaneously.

Things that cannot be clearly defined as good or evil, black or white, are what are important, and there may be things in the world that cannot be verbalized or that slip through even when one tries to explain them only with logic.

However, modern rationalists seem to think they understand things by labeling them.

'Contradiction' is a projection of high-dimensional data

Hideo Kobayashi's words, mathematically speaking, point to 'the distortion that occurs when trying to reduce high-dimensional reality into low-dimensional language (labels)'to.
.
If you project a three-dimensional object onto a two-dimensional shadow, overlaps and contradictions will inevitably occur, but modern rationalism is to dismiss this as 'it is wrong because it is contradictory,' while what Hideo Kobayashi aims for—if we were to put it in terms of a higher-order OS—is to perceive that 'this contradiction is proof of the high-dimensional truth behind it.'
.
'Not stopping thought' = 'Iteration'
The sentence 'Thinking that does not face contradictions is half-baked' is the same warning as 'Do not be satisfied with a local solution (a low mountain)'as a warning. Stopping by climbing a low mountain of simple logic (good-evil dualism) and saying 'I have found the answer' is avoiding computation, and only by continuing to run the calculation (thinking through) even when an 'error message' called contradiction appears can one reach the global optimum (the truth of the work).


'Labeling' is information loss due to compressionThe act of modern rationalists labeling things and thinking they understand them is
'irreversible compression of information.'They strip away important data as 'trash' and convert it into an easy-to-handle size (symbols). What Hideo Kobayashi is trying to protect is the 'humanity that exists precisely within that stripped-away noise.' The following code visualizes how a higher-order OS processes the depth of information that a 'labeling OS' would overlook.

Python code: Visualization of high-dimensional 'contradiction' and iteration

import numpy as np
import matplotlib.pyplot as plt

# 1. 【高次元データの生成】 
# 「人間性」や「歴史」のような、ラベル化できない複雑な真実
t = np.linspace(0, 2 * np.pi, 1000)
# 三次元空間における複雑な螺旋（真実の構造）
x = np.sin(t) * (1 + 0.5 * np.cos(10 * t))
y = np.cos(t) * (1 + 0.5 * np.cos(10 * t))
z = 0.5 * np.sin(10 * t)

# 2. 【低次元OSの処理：非可逆圧縮】
# 単純な善悪二元論（X軸のみ）に投影し、情報を削ぎ落とす
low_os_labels = np.sign(x) # 1(善)か-1(悪)かだけで判断

# 3. 【高次OSの処理：イテレーション（反復計算）】
# 矛盾を抱え込みながら、大域的最適解を探すプロセス
# ここでは「矛盾の複雑さ」を「振幅」として保持する
high_os_thought = x + y + z 

# 可視化
plt.figure(figsize=(12, 6))

# 左：低次元OSの視界（ラベル貼り）
plt.subplot(1, 2, 1)
plt.step(t, low_os_labels, color='red', label='Binary Labels (Good/Bad)')
plt.title("Low-spec OS: Irreversible Compression")
plt.ylim(-1.5, 1.5)
plt.legend()

# 右：高次OSの視界（矛盾の抱え込み）
plt.subplot(1, 2, 2)
plt.plot(t, high_os_thought, color='blue', label='High-dim Contradiction')
plt.title("High-spec OS: Iterative Deep Thinking")
plt.legend()

plt.tight_layout()
plt.show()

Increasing 'resolution'

Projection of high-dimensional data (distortion of dimensional compression)

The spiral (x, y, z) created at the beginning of the code is a 'complex truth' that cannot be fully verbalized. A low-spec OS (left graph) forcibly compresses this into np.sign(x), that is,
a binary value of 'plus or minus (good or evil).'This is the true nature of 'labeling.' In this process, most of the information is discarded as 'trash (noise),' and only a poor understanding remains.
'Iteration' that does not stop thinking
The graph on the right (higher-order OS) does not discard x, y, or z, but maintains the 'fluctuations' where they overlap as they are. In computer science, 'repeating calculations many times (iteration)' is necessary to solve complex problems. What Hideo Kobayashi calls 'thinking through' refers to this

process of not excluding the contradiction (the intersection of plus and minus) drawn by the waveform as an error, but deriving the outline of truth from its sumResistance to 'irreversible compression'

The essence of humanity (the vibration of the z-axis) exists precisely within the 'noise' that the low-spec OS discarded. When you run this code, it becomes immediately obvious how much information the red graph (low-spec) kills and how much rich information the blue graph (higher-order spec) maintains.

Why combine 'sine waves (sin/cos)'?

When expressing human 'karma' and 'contradiction,' or the 'study of life' that Hideo Kobayashi speaks of, in mathematical formulas, a straight line (y = ax) would be far too simple.

Sine waves have 'periodicity' and 'fluctuation'that repeat rising and falling.

Intent of the code

x = np.sin(t) * (1 + 0.5 * np.cos(10 * t))

Here, I am multiplying small waves that tremble finely (daily conflicts and contradictions) within a large wave (the general direction of a large life), thereby simulating 'complex human thinking that is not straightforward.'

Why did I make 'np.sign' the symbol of a 'low-spec OS'?

The function np.sign is a function that returns '1' if the numerical value is positive, and '-1' if it is negative.
No matter how complex and rich the emotion (like 0.5 or -0.8) may be, passing it through this function forcibly fixes it to either '1 (good)' or '-1 (evil)'.

The intent of the code

low_os_labels = np.sign(x)

This is the true nature of 'labeling by modern rationalists'. It expresses the arrogance of truncating all intermediate gradations (contradictions) and 'binarizing (digitizing)' information.

Why did I add 'x + y + z'?

I defined the thinking of a higher-order OS as high_os_thought = x + y + z.

Contradiction is the simultaneous existence of multiple perspectives (x-axis, y-axis, z-axis).
While a low-spec OS condemns based on a single perspective—the 'x-axis'—a higher-order OS performs the process of 'superimposing (synthesizing) the waveforms of all dimensions while they remain contradictory.' By performing 'addition,' the waveform becomes more complex, but all the original data (the truth of humanity) is preserved within it. This is
'resistance to irreversible compression of information.'

Why did I use both 'plt.step' and 'plt.plot'?

Left (low-spec): plt.step (staircase graph) visually expresses thinking that is rigid and lacks flexibility.

Right (higher-order spec): plt.plot (smooth curve) expresses the 'continuity of thought' that embraces contradictions and keeps thinking.

What this code demonstrates is that the stance of 'not fearing contradiction' corresponds to an extremely advanced process in the world of programming: 'continuing to calculate with high precision without truncating floating-point numbers (fine numerical values).'

Now, for example, since painters express things that cannot be explained in words through painting, there is no point in explaining them with words, but people like modern rationalists probably feel as if they have understood by reading the explanatory text of a painting.

Bergson emphasized this, and in 'Laughter,' he wrote:

'In short, we do not see the things themselves. In most cases, we are content to read the labels attached to them. This tendency, born of necessity, has been further emphasized by the influence of language. For language (apart from proper nouns) denotes all genres. Words, which record only the most common functions and ordinary aspects of things, intervene between things and us, and will hide their forms from our eyes, unless those forms are already hidden behind the necessity that created the words themselves.'
This is what Hideo Kobayashi wrote in 'Kotto' (Antiques):
'The word "antique" carries the weight of the history of human attachment and desire regarding objects, and a term like "ancient art" feels a bit too light. However, modern intellectuals are almost unaware of this. This is because they appreciate art but do not handle antiques. They hardly ever think about how these two preferences differ, what they have gained from art appreciation, or why they look down on handling antiques. Antiques are things to be handled; art is something to be appreciated. Saying this might sound like a meaningless pun, but it is not. The first to notice this subtle nuance were the antique dealers, and before anyone knew who started it, the term "appreciative pottery," which was unthinkable in the past, has come into common use. The term is strange, but from the antique dealer's perspective, it refers to pottery that, unfortunately, cannot be handled even if one wanted to.'
He is saying that you cannot understand antiques by just appreciating them in a museum; you have to actually touch them, which I believe also means that the essence of things cannot be understood just by looking at labels.

Since Hideo Kobayashi was strongly influenced by Bergson, it seems to me that their ways of thinking are similar, and Kobayashi was also a writer who thoroughly rejected labels.

He extended the concept of 'irreversible compression of information (labeling)' further into the physical dimension of 'resolution and interface (contact).'

What Bergson and Kobayashi both advocate is the importance of directly 'synchronizing (touching)' with the subject without going through a 'slow, low-resolution interface' called 'language (labels).'

Python code: Labels (appreciation) vs. Structure (touching antiques)

In this code, I defined the essence of things as a 'complex 3D shape' and visualized the overwhelming difference in the amount of information when viewed through 'labels (verbal explanation)' versus when handled as 'raw data (actually touching it).'

import numpy as np
import matplotlib.pyplot as plt

# 1. 【物事の本質（骨董そのもの）】
# 複雑で、触らなければわからない「多次元の質感」
t = np.linspace(0, 10, 500)
raw_essence = np.sin(t) * np.exp(0.1 * t) + np.random.normal(0, 0.2, 500) # 質感、重み、体温

# 2. 【近代合理主義者の視点（鑑賞陶器・ラベル）】
# 言葉による「非可逆なサンプリング」。情報がスッカスカになる
labels_interval = 50
label_data = np.zeros_like(raw_essence)
label_data[::labels_interval] = raw_essence[::labels_interval]

# 3. 【小林秀雄の視点（骨董いじり）】
# 実際に触れ、連続的に情報を取得し続ける
touch_data = raw_essence

# 可視化
plt.figure(figsize=(12, 6))

# 上段：ラベルを通した視界（鑑賞）
plt.subplot(2, 1, 1)
plt.stem(t, label_data, linefmt='r-', markerfmt='ro', basefmt=' ', label='Labels (Language/Genres)')
plt.title("Low-spec Interface: Reading the Tag (Information Loss)")
plt.legend()

# 下段：直接触れる視界（骨董いじり）
plt.subplot(2, 1, 2)
plt.plot(t, touch_data, color='green', label='Direct Sync (Touching the Essence)')
plt.fill_between(t, touch_data, color='green', alpha=0.2)
plt.title("High-spec Interface: Direct Interaction (Full-resolution)")
plt.legend()

plt.tight_layout()
plt.show()

Why did I add np.random.normal (noise)?

The sensation of 'touching' an antique contains a roughness and weight that cannot be put into words—in other words, 'noise (irregularity).'
Modern rationalists exclude this noise as 'waste,' but Bergson and Kobayashi realized that the 'thing itself' is hidden within this noise. By intentionally mixing in random numerical values, I expressed the 'raw reality before it is organized (labeled).'

Why did I use plt.stem (sparse points)?

The act of merely looking at the word 'art pottery' is a state where the information acquisition interval (sampling rate) is extremely low.
The vast 'blank space' between the red dots is the essence that modern rationalists overlook by 'thinking they understand.' Words are convenient, but they can only capture the essence as 'points'.

Why did I fill it in with plt.fill_between?

The act of 'touching directly' is to eliminate the boundary between the object and yourself, and to receive information continuously.
The area filled by the green graph is the 'total amount of information (intelligence)' that we are receiving, and it possesses an overwhelming data density that cannot be compared to the graph with only red dots above.

Thank you for reading this far.

The header image is from Rubens' 'The Lion Hunt' 😌