[讨论]听觉处理 auditory processing in human&a

To石悲:fficeffice" />

 

Your point 1:         根据突触的修剪理论，出生时的突触并不多，而是不断地增长，直到二岁左右（一说到一岁左右），以后则慢慢减少。这与前面的出生三个月为语言形成关键时期好象不太一致。

 

The theory of Synapse Elimination becomes effective since the moment of the baby entering this world. The growth of synapses mainly happens in the higher level functional cortexes. My hypothesis did not say “前面的出生三个月为语言形成关键时期”, instead, it said that the Auditory Cortex, particular in A1 area, can be individually defined by the very early experience of the subject. And because language sound is the most important part of that experience, the A1 area should be very sensitive to the early language experience.

 

Your point 2:         不同语言所存在的音频上的差异，如果存在的话，也就是一个非常小的范围。假如由于不同的语言造成听觉的缺失，也就是对这个特殊频率的敏感细胞活动的抑制甚至破坏，可能性几乎是零。

 

This statement is too subjective. Why it is “可能性几乎是零”?

 

Your point 3:         把语言与初级听觉区的突触和细胞发育联系在一起，是十分勉强的。但是，把语言与次级听觉区，特别是与时音相关的次级区联系起来，倒是有可能的。因为，我们听到的声音不仅仅是刺激的频率和强度，而且更是与与时间相关的声音特征。次b区的神经元网络是对初级区所带来的声音模式中的特征与时间因素联系起来分析特征的功能区。语言中所包含的声音特点，可能还要到更高级别听觉相关脑区的才能分析出来。

 

Based on the theory of developing brain, the development progress starts from the lowest level functional area to the highest level. You think”语言中所包含的声音特点，可能还要到更高级别听觉相关脑区的才能分析出来” in the first 3 months. It is another questionable statement.

 

Your point 4:         由于我们的语言能力的发展，不仅仅是被动的接触（不管是听还是看），还与我们的本身言语功能结合在一起，所以这个过程会与额叶的运动区域，还有确定思维与行为的意向选择的前额叶功能联系在一起。

 

Here is the similar issue as in Point 3.

 

Your point 1:         仅仅分析听觉区（更不用说只分析初级听觉区），要解决语言对大脑神经发育的影响是困难的。

 

I can agree with this statement. And I don’t think my hypothesis was trying to “解决语言对大脑神经发育的影响”.

引用：

---

原文由 thinker_jeff 发表于 2005-11-8 22:38:33 :

fficeffice" />

Your point 2:         不同语言所存在的音频上的差异，如果存在的话，也就是一个非常小的范围。假如由于不同的语言造成听觉的缺失，也就是对这个特殊频率的敏感细胞活动的抑制甚至破坏，可能性几乎是零。

 

This statement is too subjective. Why it is “可能性几乎是零”?

---

我认为这是个很关键的问题,它的正确与否直接影响thinker_jeff的假设是否成立.我很主观地赞成石悲的观点认为"可能性几乎为零".

thinker_jeff假设是:日本人婴儿(0-3个月)的语音环境里缺失某些频率,所以3月龄之后的A1对这个缺失的频率不敏感.我之所以反对是因为语音环境只是整体声音环境的一部分,语音环境里缺失某些频率应该可以被整体声音环境补充上的.语音中不包含800HZ-20KHZ的频率,但人一样听得到,这就是例证.

我们也可以探讨是否可以通过实验来证明它,我想通过观察日本人subjects听"热"和"乐"的mmn的颞成分就可以证明石悲的话是正确还是错误的.我就在日本,可以为大家免费提供实验环境,可以考虑做共同研究.

---

 

Your point 3:         把语言与初级听觉区的突触和细胞发育联系在一起，是十分勉强的。但是，把语言与次级听觉区，特别是与时音相关的次级区联系起来，倒是有可能的。因为，我们听到的声音不仅仅是刺激的频率和强度，而且更是与与时间相关的声音特征。次级区的神经元网络是对初级区所带来的声音模式中的特征与时间因素联系起来分析特征的功能区。语言中所包含的声音特点，可能还要到更高级别听觉相关脑区的才能分析出来。

 

Based on the theory of developing brain, the development progress starts from the lowest level functional area to the highest level. You think”语言中所包含的声音特点，可能还要到更高级别听觉相关脑区的才能分析出来” in the first 3 months. It is another questionable statement.

---

我认为,A1的 role是检测,既发现有没有声音;A2的role是辨别识别.有关声音的辨别识别的研究并不多,但从已经知道的结果中我们可以发现它并不简单.声音的辨别识别甚至受到其他的感觉输入影响,如"听幻觉".我主观地认为日本人的"R""L"不分是"听幻觉"的一种.因为我在日本的5年多生活中发现很多人可以熟练地分清"R""L",即使普通日本人,教他发音技巧,也可以说出"L"的.如果大家感兴趣,我可以开一个关于"听幻觉"的帖子,那也是我的研究课题之一.

［此帖子已被 zhixl 在 2005-11-9 11:26:31 编辑过］

To zhixl:fficeffice" />

 

            I am so glad you told us, “我就在日本,可以为大家免费提供实验环境,可以考虑做共同研究.” Some discussion in this BBS can become experimental research that is over my expectation. No matter who is right or wrong, we can gain the real knowledge in this field. Thank you so much for this offer.

 

I am very surprise that you were saying “我认为, A1的 role是检测, 既发现有没有声音; A2的role是辨别识别.” Either my knowledge is incorrect or your statement is inaccurate. You know we are not arguing about the A2 area so let’s focus on A1 area. The A1 area is a short name of primary auditory cortex, which I am able to find a graphic illustration Figure 11.29 in Neuroscience: Exploring the Brain (探索脑, 神经科学图片库) on this website. The role of A1 has two parts – one is mapping the auditory frequencies, which serves as the frequency discrimination center; another is detecting the amplitude of each frequency by the firing rate at the synapses. That why I believe for the new born baby those auditory inputs of “r” and “l” in English should not be projected into the exact synapses in A1. I think this is the key point causing the different views between you and me. If you can let me know what is wrong in the above understanding, it will be my gain in this discussion.

 

            If I am interpreting right, you think the different frequent features in different languages can be compensated by other auditory sources. I think this has a time issue here. The fact is there, that those Japanese babies have lost their some discriminating ability after 3 months, which has approved that in those 3 months other auditory sources did not compensate that difference. After they grown up, they can learn how to compensate the difference because the learning will make some new synapses in somewhere, where may not repeat the same places as the lost synapses. The case of learning English as second language after 7 years old is a good example.

 

            Let’s make some conclusions sooner. Then we can design the experiments we want to do. I am enjoying this topic.

下图是鼠的初级听觉皮质.

A比较清晰地标出了频率的分布.

B说明了频率和强度的相关

 

A: representative CF map of primary auditory cortex from a naı¨ve adult rat. Each polygon represents 1 electrode penetration. Color represents each site’s best frequency. Nonresponsive and auditory responding non-A1 sites are marked with dots and x’s, respectively. Scale bar 5 250 mm. B: tuning curve tips for the rat in A. The tip of each V depicts minimum threshold for each site. Width of the V represents tuning curve width 10 dB above threshold. C: representative example of the temporal selectivity of A1 neurons to tone trains. Most neurons in rat A1 do not respond to tones presented at rates exceeding 12 pulses/s. The dot raster indicates the responses to 12 repetitions of 16 rates. The short horizontal lines designate the spike collection windows used to generate the RRTF. Tone onsets occur 8 ms before each line.

To thinker_jeff fficeffice" /> 

---

 I am very surprise that you were saying “我认为, A1的 role是检测, 既发现有没有声音; A2的role是辨别识别.” Either my knowledge is incorrect or your statement is inaccurate. You know we are not arguing about the A2 area so let’s focus on A1 area. The A1 area is a short name of primary auditory cortex, which I am able to find a graphic illustration Figure 11.29 in Neuroscience: Exploring the Brain (探索脑, 神经科学图片库) on this website. The role of A1 has two parts – one is mapping the auditory frequencies, which serves as the frequency discrimination center; another is detecting the amplitude of each frequency by the firing rate at the synapses.

我所说的A1的role与thinker_jeff采用的引文并不矛盾,只是我没有清楚地表达出我的意思.这里,我对我想用"检测"和"辨别识别"表达的意思做一个相对详细的解释

检测:检测声音是否存在,是外源性的,是大脑对刺激出现的强制性反应.它只是启动了内部触发机制,但它不参与刺激的特征分析,并不能形成对该刺激的有意识的知觉(conscious perception),是被动的知觉.它不能指明刺激的内容,不能辨别刺激的特征,但它能让大脑侦知到刺激的开始(和终止).比如task 要求subjects只是听声音提示,然后回答听到没有.那么完成这个任务时,只需要"检测"就可以了,不需要"辨别识别",因为task并不要求subjects回答声音的大小位置等等.

辨别识别:是建立在"检测"之上的,对刺激的特征分析,能形成对该刺激的有意识的知觉(conscious perception),是形成主动的知觉.如要求subjects完成的task包括回答声音的大小位置等等.区分"R"和"L"也属于"辨别识别",是主动的知觉任务.

我强调A1的role是"启动了内部触发机制,但它不参与刺激的特征分析,并不能形成对该刺激的主动的知觉". thinker_jeff采用的引文指出A1的不同神经元团对不同的频率(及强度)特异反应. 我赞同某些学者的观点既A1的频率(及强度)特异分布并不代表它参与了刺激的特征分析并形成对该刺激的主动的知觉.A1的频率(及强度)特异分布的作用是启动了内部触,而这种触发具有频率(及强度)特异性,使更高级的皮质去进行特征分析(既辨别识别).

 

---

 

That why I believe for the new born baby those auditory inputs of “r” and “l” in English should not be projected into the exact synapses in A1. I think this is the key point causing the different views between you and me. If you can let me know what is wrong in the above understanding, it will be my gain in this discussion.

 这里我再强调一下我的观点,A1的作用是"听到了有声音刺激",至于听出来是"R"还是"L"则不是在A1中处理的结果.

我认为A1可以忠实地对所有阈值上的20-20KHZ的声音刺激做应答.或许在人的发育过程中这个带宽会有变化,但它不应该是针对某个特异的音源(如语音).能否分辨出"R"和"L"是知觉过程,是大脑的主动过程,是A1之后的某个step发生的.这一点和thinker_jeff观点是相左的,所以邀请thinker_jeff希望通过做一个实际的实验去检验哪个观点正确.

---

  

            If I am interpreting right, you think the different frequent features in different languages can be compensated by other auditory sources. I think this has a time issue here. The fact is there, that those Japanese babies have lost their some discriminating ability after 3 months, which has approved that in those 3 months other auditory sources did not compensate that difference. After they grown up, they can learn how to compensate the difference because the learning will make some new synapses in somewhere, where may not repeat the same places as the lost synapses. The case of learning English as second language after 7 years old is a good example.

 

你阐述的观点很有趣,不知道是不是你首次提出的,如果它正确,可以解释人类A1差异为什么比猴的A1差异要大.我主观认为同一环境下成长起来的人,A1的差异也很大,用你的假说就很难解释了?我们是否需要先做一个研究,看看成长环境(如语言环境)对A1差异的影响?

---

 

            Let’s make some conclusions sooner. Then we can design the experiments we want to do. I am enjoying this topic.

 

现在我们的观点有如下矛盾的地方,算是一个总结

我认为A1在人的一生中都是可以全频域(20-20KHZ)应答,即使应答的带宽有改变,这种改变也不具有针对音源(如语音)的特异性;A1不参与声音刺激的特征分析(尽管它的分布是频率/强度特异的).

thinker_jeff认为,A1的应答频域带宽在人的发育过程中是变化的,而且受语音环境影响很大.A1的这种变化直接影响了脑对声音刺激的特征分析(如辨别"R"和"L")

 

关于总结,如果你没什么意见我们可以讨论实验设计及需要搜集哪些先行研究的资料.

 

［此帖子已被 zhixl 在 2005-11-11 12:54:18 编辑过］

It seems Zhixl has not finished his answers today, which is very understandable. Every body should take care of the daily life first. I am waiting for the answers of other parts.fficeffice" />

 

I feel good that my knowledge about A1 was basically correct although it was not as detail and professional as Zhixl’s.

［此帖子已被 thinker_jeff 在 2005-11-10 21:34:53 编辑过］

引用：

---

原文由 zhixl 发表于 2005-11-10 14:40:36 :

现在我们的观点有如下矛盾的地方,算是一个总结:

我认为A1在人的一生中都是可以全频域(20-20KHZ)应答,即使应答的带宽有改变,这种改变也不具有针对音源(如语音)的特异性;A1不参与声音刺激的特征分析(尽管它的分布是频率/强度特异的).

 

thinker_jeff 认为, A1的应答频域带宽在人的发育过程中是变化的,而且受语音环境影响很大.A1的这种变化直接影响了脑对声音刺激的特征分析(如辨别"R"和"L").

---

Very good conclusion!fficeffice" />

引用：

---

原文由 zhixl 发表于 2005-11-10 14:40:36 :

To thinker_jeff fficeffice" /> 

---

  

            If I am interpreting right, you think the different frequent features in different languages can be compensated by other auditory sources. I think this has a time issue here. The fact is there, that those Japanese babies have lost their some discriminating ability after 3 months, which has approved that in those 3 months other auditory sources did not compensate that difference. After they grown up, they can learn how to compensate the difference because the learning will make some new synapses in somewhere, where may not repeat the same places as the lost synapses. The case of learning English as second language after 7 years old is a good example.

 

你阐述的观点很有趣,不知道是不是你首次提出的,如果它正确,可以解释人类A1差异为什么比猴的A1差异要大.我主观认为同一环境下成长起来的人,A1的差异也很大,用你的假说就很难解释了?我们是否需要先做一个研究,看看成长环境(如语言环境)对A1差异的影响?

---

I tell you the truth, at this moment I am not sure if this was an idea of somebody else which I had read in some book or I combined the information from multiple sources and put together by myself. It’s not funny when you read so much and think so much. However, I don’t know anyone had done a real research about the point in my hypothesis.

 

I suggest that our first step is searching the related information. Do you agree with me on this?

---

  I suggest that our first step is searching the related information. Do you agree with me on this?fficeffice" />

---

支持。如果能找到人类的对听觉皮质的top-down投射就好了。灵长类的也行，实在没有的话，哺乳动物也可以。有了对a1的top-down的证据，这个证据就站得住脚了。需要下载论文的时候，把连接告诉我，我从我所在的研究机构下载，然后传给你。

This is a good one!

_______________________________________________________________

Abstract #5708, Date Sunday, Feb 20 2000 1:00PM - 12:00PM , Session ,

Frequency tuning plasticity in rat primary auditory cortex

Pritesh K Pandya , Jessica Vazquez , Daniel Rathbun , Navzer D Engineer , Raluca Moucha , Michael P Kilgard

It is well documented that powerful plasticity mechanisms can substantially alter the receptive fields (RF) of auditory cortex neurons. Episodic electrical stimulation of the cholinergic nucleus basalis (NB) paired with tonal stimuli can be used to mimic naturally occurring experience-dependent plasticity. This plasticity paradigm is sufficient to generate dramatic changes in the representation of both spectral (Kilgard & Merzenich, Science, 1998) and temporal (Kilgard & Merzenich, Nature Neuroscience, 1998) information in the primary auditory cortex (A1) of adult rats. RF size (spectral selectivity) can be narrowed, broadened, or left unchanged depending on the specific parameters of the stimulus paired with NB activation. Previous work using NB stimulation suggests that frequency bandwidth varies systematically as a function of spectral variability and modulation rate. A primary goal of our ongoing research is to precisely characterize how differential acoustic experience shapes cortical response properties. The present study represents a further step in our effort to determine the rules of plasticity for frequency selectivity. Adult rats were chronically implanted with NB stimulating electrodes and received electrical stimulation of NB paired with a single tone with a high degree of spectral variability (1 of 10 randomly interleaved tone frequencies) 200-300 times per day for approximately 20 days. Frequency-intensity tuning properties were derived using multi-unit data from 50-70 microelectrode penetrations for each animal during an acute mapping experiment using barbiturate anesthesia. Our present model of A1 plasticity predicts that pairing NB stimulation with different tone frequencies will trigger maximal sharpening of A1 frequency tuning. Preliminary results are encouraging and the final results of this study will be discussed in the context of previous work on the differential RF plasticity in A1 dependent upon the degree of spectral variability and modulation rate of the paired stimuli.

Supported by the Callier Excellence in Education Fund

［此帖子已被 thinker_jeff 在 2005-11-12 22:59:41 编辑过］